JT9D Engine Thrust Estimation and Model Sensitivity Analysis Using Gradient Boosting Regression Method

Round 1

Reviewer 1 Report

The research is of a very high standard. Several issues should be improved to increase the quality of the paper.

1) All physical quantities used in formulas should be explained below. Without such a description, it is hard to verify and understand them.

2) Please explain why those 12 features were selected as inputs (line 136).

3) Parameters HPC-Pt or LPC-Pt are listed in Table 2. In the text, different designation is used, i.e. only HPC and PLC, while for other parameters named in Table 2, complete designations are used in the text. Please, standardize it.

4) Fonts used to describe Figure 3 could be bigger.

5) Please support why values 0.8 (for colsample_bytree) and 0.5 (training sample's subsample ratio) were used in tests (lines 183 and 184). How do those assumptions influence the results?

6) In discussion, the higher deviation of obtained results is explained because of the smaller test dataset. Can the Authors name the amount of test dataset recommended not to get too high a deviation value? It would be valuable information for future users of the Authors' method.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

This paper introduces an application of XGBoost to estimate the thrust of a JT9D engine. They also present a thorough sensitivity analysis, providing optimal hyperparameter values.

The most significant concern regarding this manuscript is the apparent lack of a clearly defined novelty in the presented work. While the application of the XGBoost model and the subsequent hyperparameter tuning are conducted, these methods are well-established within the field. The XGBoost model, although a solid and well-established method, is basic in the current, rapidly evolving field of machine learning. It could be beneficial to explore the application of more advanced or cutting-edge predictive models to enhance the innovative contribution of the paper.

Another concern is the validation of the simulation.

1. Introduction:

The introduction could be expanded to provide a clearer picture of the current state of research in the field, and how their work fits into it. The authors have not provided a clear motivation for the current study. Why is it important to improve thrust estimation for the JT9D engine? What are the limitations or inaccuracies in the current estimation methods that their approach might address?

A few minor grammar errors are present that should be fixed. For example, "The NASA" should be "NASA," and "parameter's effect" should be "parameters' effects."

The authors should explain how understanding the operation of the JT9D engine and the principles of parametric cycle analysis is important for their study.

2. Materials and Methods

The authors need to explain how their simulations were validated. Do they have real-world data against which the simulations can be compared? If not, they need to discuss the limitations this might impose on their study and how they plan to address them. If possible, the authors should compare their simulation results with those of other similar studies. This could provide additional validation if direct comparison with real-world data is not possible.

While the authors have given a brief explanation of what violin plots represent, they haven't used them to convey any specific insights about the data they have collected. The authors should provide a more detailed interpretation of the violin plots.

4. Discussion

The fact that the model's predictions deviate more significantly at higher engine thrust values might suggest some limitations of the XGBoost model when dealing with this particular problem.

N.A

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

the manuscript has been sufficiently improved to warrant publication in Aerospace