Predicting the Recovery and Nonrecoverable Compliance Behaviour of Asphalt Binders Using Artificial Neural Networks
Round 1
Reviewer 1 Report
Dear Author.Congratulations for the work done.
The manuscript proposes the use of ANN for the prediction of recovery and Nonrecoverable compliance behaviour of asphalt binders. The subject of the article is very interesting in the field. The research methodology is reasonable, the research is comprehensive and some very interesting and innovative conclusions are obtained.
In general the structure and content of the manuscript is acceptable for the Processes. In order to improve its readability, please consider these suggestions out of the following.
1 What are the criteria for the selection of additive dosing and why the dosing of various modifiers varies
2 MSCR grade is based on that index to assess?
3 Line 309: What is the modifier dose for various types of modified asphalt aging
4 The conclusion is too simple and only shows that it is feasible to use ANN to make predictions. In fact, before this modified asphalt performance with time and frequency of the test results can also be summarized as the study conclusions.
5 The effect of different modifiers on the recovery and nonrecoverable compliance behaviour of asphalt can also be compared and analyzed.
6 The paper only studied the recovery and Nonrecoverable compliance behaviour of asphalt binder, the powder to rubber ratio also affects the performance of asphalt mastic, which can be studied in depth.
Author Response
Please see the attachment
Author Response File: 
Author Response.pdf
Reviewer 2 Report
In this manuscript, the authors aimed to investigate the possibility of using the artificial neural network (ANNs) model to predict the non-recoverable compliance and recovery properties of modified asphalt binders. Over 880 data points were used using eleven mixed asphalt binders. Frequency sweep and multiple stress creep-recovery (MSCR) tests were conducted on all binders at different temperatures.
The literature is well-described, making readers understand a brief studying history of this field.
The research methods are well-described, giving readers sufficient to follow the study tools. Jnr and R of unaged and aged asphalt binders were predicted using an ANNs model which was developed using five inputs: test temperature, frequency, storage modulus, loss modulus, and viscosity. The training set often contains the primary data, while the validation and test sets typically contain a smaller part of the data. In this study, 70% of the total data was applied to train the ANNs, 15% was used to validate the neural network, and 15% was utilized for testing the capability of the ANNs to predict the output.
It’s a good thing to see that complex problems are tried to be solved by machine learning or deep learning tools.
In summary, the manuscript can be published after minor revision.
Some minor suggestions are listed below.
-----------------------------------------------
-----------------------------------------------
[[Minor suggestions]]
The suggestions are structured as shown below.
[Suggested point][Position]
Descriptions.
1. [Abbreviation][Line 25]
The abbreviation (HMA) should be stated with full names before using it.
2. [Terminology description][Line 49]
Brief descriptions or citations are suggested on the term “French rutting test.”
3. [Typo][Line 89]
It is ANNs instead of ANNS.
4. [Abbreviation][Line 100]
The abbreviation (DSR) should be stated with full names before using it.
5. [Abbreviation][Line 123]
The abbreviation (DSR) has been stated with full names and can be used directly here.
To deal with such problems. Please check the entire manuscript.
6. [Abbreviation][Line 132]
The abbreviation (ANNs) has been stated with full names (line 54) and can be used directly here.
7. [Abbreviation][Line 142]
The abbreviation (SBS) has been stated with full names (line 11) and can be used directly here.
8. [Terminology][line 145]
The term “ASTM C618-17a” should be introduced briefly. For example, if it is an instrument (or method), the sentence could be “…according to the instrument (or method) ASTM C618-17a.”
9. [Abbreviation][Line 158]
The abbreviation (DSR) has been stated with full names (line 100) and can be used directly here.
10. [Typo][Line 164][Line 246]
The unit of frequency is “Hz.” The “z” is written in normal font instead of subscript.
11. [Figure 8, 10~15]
The resolution of the figures is too low, please increase the resolution for better reading.
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Reviewer 3 Report
The study use Artificial Neural Networks (ANNs) to predict the recovery (R) and nonrecoverable compliance. The materail method section of this study is very excellent. Each details are presented in good way. However, there is lack of content in introduction and results section. The reviewer strongtly suggest that the authers complete each comment of the reviewer.
In the abstract, give some results based on %
Novelty is not clear. . Please write difference between prevous studies and make a clear statement about novelty.
A general pragraph should be added to introduction for the importane of using waste materials to solve some enviromental problems since the authors utilized waste materials and also geopolymer which is eco-friendly materials. Following studies must be used for this pragraph: use of recycled coal bottom ash in reinforced concrete beams as replacement for aggregate; flexural behavior of reinforced concrete beams using waste marble powder towards application of sustainable concrete; improvement in bending performance of reinforced concrete beams produced with waste lathe scraps; performance assessment of fiber-reinforced concrete produced with waste lathe fibers; performance evaluation of fiber-reinforced concretes produced with steel fibers extracted from waste tire; performance evaluation of fiber-reinforced concretes produced with steel fibers extracted from waste tire; composition component influence on concrete properties with the additive of rubber tree seed shells; normal-weight concrete with improved stress–strain characteristics reinforced with dispersed coconut fibers; effects of waste powder, fine and coarse marble aggregates on concrete compressive strength
Material method section should be impvored.
More comments should be provided for Tables 1 and 2
No information was provided on Geopolymer in introduction. Give a summary for this.
No information was provided for both glass powder and fly ash. Please give details for this and use followings influence of replacing cement with waste glass on mechanical properties of concrete; concrete containing waste glass as an environmentally friendly aggregate: a review on fresh and mechanical characteristics; mechanical behavior of crushed waste glass as replacement of aggregates
Add future needs for conclusion
Add recommendations based on your findings.
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
The manuscript proposes the use of ANN to predict the restorative and irrecoverable compliance behavior of asphalt binders. The subject of the article is very interesting in the field. The research methodology is sound and comprehensive and leads to some very interesting and innovative conclusions. The revised article can be accepted.
Author Response
-
Reviewer 3 Report
The reviewer did not see any improvements in the updated file.
Please indicate detail response for 1 st revision.
Highlight your text.
Author Response
-
