A Data-Driven Framework for Early-Stage Fatigue Damage Detection in Aluminum Alloys Using Ultrasonic Sensors
Round 1
Reviewer 1 Report
The article concerns important and current issues in the field of assessing the condition of machine and device elements. It is a dynamically developing field, which implies the topicality of the theatics undertaken by the authors.
The article does not contain significant content-related flaws, however, several issues listed below require a more detailed explanation.
- Why was such an ultrasonic sensor chosen? Have there been any preliminary studies on the effectiveness of using sensors of other frequencies?
- There is a lack of information on the methodology used in the proposed algorithms:
- learning methods,
- the size of the training data set,
- learning accuracy (verification based on known test data),
- structure of the block classifying the input data. - In this type of problem, parallel computing with the use of GPU is often used. Have the authors considered the possibility of using this technology for calculations?
- There are some minor editing errors, such as "Error! Reference source not found" errors, which must be removed.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
This manuscript presents machine learning approach to detect fatigue damage in notched specimens. In my opinion the manuscript is well written and the results are presented clearly. I noticed several formatting errors with links to the references. Please fix those. Besides that I don't have any further questions to authors.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
