Round  1

Reviewer 1 Report

This paper deals with TFM methods for thin walled structure inspections.

after the reading of this papers, I had several remarks :

1/  The TFM method is quite popular, as the FDTD methods, but the authors give few references on these topics. In general, references has to be added to the paper.

2/ the defect positions and sizes should be added to figures

3/ the figure 13 shows a huge difference between theoretical and measured size of default but the authors say that the methods allow to measure default longer than 1 mm. Moreover the authors said that SS-SS modes are the best for circular defaults but any results are shown.

4/ stir friction welding should lead to anisotropy. How the authors deal with that?

5/ the authors said that multi mode imaging are needed to correct testing results for sample with multi defaults. Could they add that in the results?

Author Response

To reviewer

Thank you for your advices on all problems of my manuscript coatings-440045. After nearly 10 days’ efforts, most parts of the article have been revised according to your advices. However, due to deadline of coating, I have to upload the revise without the detail discussion of experimental results using indicator DSNR for each embedded defect in accordance with the analysis pattern of section 3.

Q1: The TFM method is quite popular, as the FDTD methods, but the authors give few references on these topics. In general, references has to be added to the paper.

A: 15 more references related to the article have been added. Among 11 of them are related to ultrasonic NDT.

Q2: the defect positions and sizes should be added to figures.

A: The reason why no defect information is added to figures is that there exist too many defects of different types, sizes and locations. It’s very redundant to describe all of them using figures. So, the location and size of each defect is defined by size and base point given in Table. 2. The co-ordinate system used to define defect location can be seen in Fig. 4(a) or 4(b). The meaning of base point is given in Fig. 4(c).

Q3: the figure 13 shows a huge difference between theoretical and measured size of default but the authors say that the methods allow to measure default longer than 1 mm. Moreover the authors said that SS-SS modes are the best for circular defaults but any results are shown.

A: A new indicator named Defect SNR (DSNR) has been introduced in this paper. The definition is given in E.q. (7). All analyses of imaging performance of different modes are re-written using DSNR, mean of DSNR and the standard deviation of DSNR. They are given in Table. 3 to 6. Based on DSNR, it’s obvious to find that SS-SS modes are the best for circular defaults. Table. 7 has been added to give the actual size of defects and estimated size with wedge coupling. Besides, the linear sizing equation for lower half region rectangular defect using linear regression are given at the same time.

Q4: stir friction welding should lead to anisotropy. How the authors deal with that?

A: The anisotropy is not considered in this paper as the main topic of this paper is about thin-walled structure ultrasonic array inspection. It’s just a coincidence that our available thin-walled samples with different types of defects are manufactured by stir friction. However, the group velocity variation due to anisotropic may have some contribution to the SNR decrement in experimental TFM. The above discussion has been added into section.

Q5: the authors said that multi mode imaging are needed to correct testing results for sample with multi defaults. Could they add that in the results?

A: Multi mode imaging are needed to correct testing results for sample with multi defaults means that the combination of SS-SS, SS-S, SSS-SS mode TFM imaging results is not enough to give a correct testing result. For sample with more complicated defect distribution, other modes with higher reflection times like SSSS-SSS, SSSS-SSSS should be combined to improve the inspecting result. This statement has been added into the conclusion section.

Reviewer 2 Report

Comments for author:

The manuscript deals with the inspection of thin-walled structures by NDT using ultrasonic array. The authors employed the theoretical and experimental analysis to compute the SNR and size of defects. This manuscript is poorly written and sometimes it misleads the technical definition of the sentences. Authors are requested to take help of native English speaker or look for the proofreading facility by the publisher. There appear to be critical problems in the submitted research results and lack of comparison with theoretical/numerical and experimental results. Therefore, it is highly recommended that authors should majorly revise the manuscript after accounting for the following comments below, before considering to be published in the journal of “Coating”.

1. Please check the title of the article. Are you analyzing the defects/holes or methods? The title should be related to the aim of research. Your aim seems to estimate the size of defects in thin-walled structures by ultrasonic array or the possibility of estimation.

2. In the abstract:

 Please rewrite the abstract by stating the problem.

 Please mention the novelty you bring in this article.

 What does it mean theoretically if you performed numerical simulations?

 Again, what is the accuracy of detection and estimation (if you performed) of the defects. You must include this to demonstrate why your method is novel or accurate.

3. In Section 1 (Introduction):

 Again, please check the sentence structures properly throughout the text.

 Authors are suggested to include the latest references. There are a lot of works already performed based on ultrasonic array and ultrasonic NDT of structures. However, only 2 references are cited in the introduction of the manuscript which clearly shows that the review of literature is not performed well. Please go through the literature and cite at least 5 to 10 more references related to ultrasonic NDT.

Here are some examples of references:

[1] R. Raišutis, O. Tumšys, and R. Kažys, “Development of the technique for independent dual focusing of contact type ultrasonic phased array transducer in two orthogonal planes,” NDT & E International, vol. 88, pp. 71–80, Jun. 2017.

[2] Drinkwater BW, Wilcox PD. Ultrasonic arrays for non-destructive evaluation: a review. NDTE Int 2006;39(7):525–41.

[3] Duxbury D, Russell J, Lowe M. The effect of variation in phased array element performance for non-destructive evaluation (NDE). Ultrasonics 2013;53(6):1065–78.

[4] Shih, J.-L.; Wu, K.-T.; Jen, C.-K.; Chiu, C.-H.; Tzeng, J.-C.; Liaw, J.-W. Applications of Flexible Ultrasonic Transducer Array for Defect Detection at 150 °C. Sensors 2013, 13, 975-983

 How your technique is novel as compared to the research already performed?

 Authors should clarify why ultrasonic testing is the promising technique as compared to other NDT techniques. (Cite some latest articles as well).

 The aim of the research should be clearly explained. What parameters (SNR or size of defects and what type of defects and their location) you have to estimate?

4. In Section 2 (Linear ultrasonic array method for general thin-walled structure):

 Between Line 40-50:

---“Suppose the longitudinal and shear velocity of general thin-walled structure...?”

--- velocity of propagating wave or velocity of structure?

---what are x and z? Please denote each symbol you use.

---What is j: 0 to N or 1 to N?

---Line 50: after Hilbert transformations (where are the reference and citation?)

 Between Line 51-60:

---In figure 1, which one is Layer 0 and layer n, it is not clear. Please redraw the figures with mentioning the layer notation?

---Again Ci is not the velocity of the layer. Please correct it.

---“TOF can be solved numerically using Snell’s law or..” (where are the reference and citation?)

---- Clearly explain Fig. 1a and 1b. Moreover, redraw them to understand for the readers.

 Between Line 60-100:

---- Section 2.1 and 2.2: Are all equations developed by the authors for the explanation of phased array and total focusing method? If not, cite them properly. --- The time of flight is T or TOF as used in Line 56.

 Eq. 1-Eq. 4 are not cited. Are they developed by the authors themselves? If not, proper citations are required.

5. In Section 3 (Theoretical analysis):

 Again, please check the sentence structures properly throughout the text.

 Please correct throughout the text that the velocity of propagating waves (Shear /longitudinal) in Aluminium instead of the velocity of Aluminium.

 Please mention the SNR value of each Image if you are comparing the SNRs.

 Please prepare a Table to compare the actual size of defects and estimated size with wedge coupling.

6. In Section 4 (Experimental analysis):

 Again, please check the sentence structures properly throughout the text.

 Please add the photo of object and array of transducers in Figure 14.

 Compare the experimental and numerical results in terms of defect size with accuracy in measurement.

7. In Section 5 (Conclusion):

 The conclusion section should be fully rewritten to show what the main input of this work to this area in a concise way.

Finally, there are lots of variables and abbreviations in this manuscript, a Nomenclature Section should be added.

Author Response

-1. To Reviewer

Thank you for your advices on all problems of my manuscript coatings-440045. After nearly 10 days’ efforts, most parts of the article have been revised according to your advices. However, due to deadline of coating, I have to upload the revise with the detail discussion of experimental results using indicator DSNR for each embedded defect in accordance with the analysis pattern of section 3.

0. Revision of sentence structures

Q0: This manuscript is poorly written and sometimes it misleads the technical definition of the sentences. please check the sentence structures properly throughout the text

A: The sentences throughout the text has been revised in my best efforts.

1. Revision of title

Q1: Are you analyzing the defects/holes or methods? The title should be related to the aim of research. Your aim seems to estimate the size of defects in thin-walled structures by ultrasonic array or the possibility of estimation.

A: The aim of this paper is indeed to analyze the image result of thin-walled structure whose thickness is 5 times less than the longitudinal wave-length so that to understand which imaging mode and sizing method are preferable to specific type of defect.

2. Revision of abstraction

Q2.1: Please rewrite the abstract by stating the problem.

A: The abstract has been rewritten. The reason why the inspection of structure with thickness 5 times smaller than the acoustic wave-length is considered in this paper has been stated.

Q2.2: Please mention the novelty you bring in this article.

A: The novelty of this article includes two things. First, linear ultrasonic array methods using different combinations of wave mode, reflection time and coupling condition are combined to inspect the thin-walled structure. Second, the thickness is 5 times smaller than the acoustic wave-length.

Q2.3: What does it mean theoretically if you performed numerical simulations?

A: Yes, the numerical simulations full matrix is used to analyze the theoretically performance of ultrasonic array methods.

Q2.4: Again, what is the accuracy of detection and estimation (if you performed) of the defects. You must include this to demonstrate why your method is novel or accurate.

A: For simulate circular defect, the diameter can be measured according to the maximum image amplitude of defect. For simulate rectangular defect located in lower half region, the nominal length can be measured using a linear function whose input is the -6dB drop length of SS-S mode image.

3. Revision of section 1 (Introduction)

Q3.1: Authors are suggested to include the latest references. There are a lot of works already performed based on ultrasonic array and ultrasonic NDT of structures. However, only 2 references are cited in the introduction of the manuscript which clearly shows that the review of literature is not performed well. Please go through the literature and cite at least 5 to 10 more references related to ultrasonic NDT.

A: 15 more references related to the article have been added. Among 11 of them are related to ultrasonic NDT.

Q3.2: How your technique is novel as compared to the research already performed?

A: New sentences “Current researches on thin-walled structure inspection are mainly conducted by array transducers with high central frequency. For 6mm thick titanium thin-walled structure with 6000m/s longitudinal wave velocity, directly contact testing and evaluation using array transducers whose central frequency is larger than 10MHz have been detailly analyzed. In this case, the thickness of structure is 10 times larger than the longitudinal wave-length. However, there are few researches discussing about the situation when the thickness to wavelength ratio is much smaller than 10. Unfortunately, due to the manufacturing technology and cost restriction, the thicknesses of many thin-walled structure will be 10 times 40 smaller than the minimum longitudinal wave-length of available array transducer.” from line 33 to 40 has added in front of sentence “The aim of …” to describe the novelty of technique in this paper.

Q3.3: Authors should clarify why ultrasonic testing is the promising technique as compared to other NDT techniques. (Cite some latest articles as well).

A: The old sentence “Ultrasonic is one of the most widely used NDT&E approaches in industry.” has been replaced by new sentences “To monitor the integrity and status of components during their total life cycle, appropriate non-destructive testing and evaluation (NDT&E) methods are required to find the potentially existing interior flaws such as holes and cracks. NDT&E methods like infrared is very sensitive to surface defects and can be applied to out-field without any pollution. However, it's not sensitive to inner defects. CT has very high imaging resolution for volume type defects like holes. However, it's not sensitive to 2-demensional discontinuities like cracks and the radiations are harmful to human body. Ultrasonic is one of the most promising NDT&E methods without any pollution and harm. Defects inside sample can be located and sized directly using the A-Scan signal received by single element ultrasonic transducer. It's very sensitive to both volume and two-dimensional defects in elastic material.” from line 4 to 14 to describe the reason why ultrasonic testing is promising technique as compared to other NDT techniques.

Q3.4: The aim of the research should be clearly explained. What parameters (SNR or size of defects and what type of defects and their location) you have to estimate?

A: The old sentences “These plate structures are embedded with defects of different geometries to analyze their theoretical performances. The thickness satisfying the definition of thin-walled as the longitudinal wavelength of 5MHz acoustic wave in Aluminum is 1.3mm.” have been replaced by new sentences “These plates are embedded with defects of different geometrical types. Defect location, maximum defect image amplitude, defect sizing ability and defect SNR (DSNR) are used to theoretically analyze their performances.” from line 45 to 57 to introduce the parameters used in this paper to analysis the array imaging method.

Revision of section 2 (Linear ultrasonic array method for general thin-walled structure)

Q4.1: Suppose the longitudinal and shear velocity of general thin-walled structure...?

A: The wrong term “the longitudinal and shear velocity of general thin-walled structure...” and “The velocity of coupling material…” have been replaced by “the longitudinal and shear velocity of acoustic wave in thin-walled structure...” and “The velocity of longitudinal wave in coupling material…” In Table. 1.

Q4.2: what are x and z? Please denote each symbol you use

A: The symbols x and z have been added into Fig. 1.

Q4.3: What is j: 0 to N or 1 to N?

A: The old sentence “… all combinations of transmit and receive elements i and j.” has been revised “… all transmit and receive elements combinations.” in line 54. On the other hand, the symbol M_ij(t) of full matrix has been revised to {M_ij(t) | i, j = 1, 2, …, N} in line 56 to explain meanings of i and j.

Q4.4: Line 50: after Hilbert transformations (where are the reference and citation?).

A: The usage of Hilbert transform before full matrix processing is a trivial approach in ultrasonic array testing and evaluation researches. More details about Hilbert transformation and its application can be found in reference [9] (E.q. (8) at the end of section 2).

Q4.5: In figure 1, which one is Layer 0 and layer n, it is not clear. Please redraw the figures with mentioning the layer notation?

A: The symbols layer 0 to n have been added into Fig. 1.

Q4.6: Again Ci is not the velocity of the layer. Please correct it.

A: All wrong term “velocity of the layer” in this paper have all been replaced by “wave velocity in layer”.

Q4.7: TOF can be solved numerically using Snell’s law or..” (where are the reference and citation?

A: The function TOF_n is used to solve the travelling time of acoustic ray in multi-layered medium. The calculation using Snell’s law or Fermat minimum principle is very trivial in ultrasonic array testing and evaluation researches. More details about the calculation based on them can be found in reference [10, 11].

Q4.8: Clearly explain Fig. 1a and 1b. Moreover, redraw them to understand for the readers.

A: The Fig. 1(a) and Fig. 1(b) have been completely redrawn. The reflection and wave mode of Fig. 1(a) and Fig. 1(b) is used in Eq. (2) to illustrate the meaning of parameters of function TOF_n and how reflection and wave mode are used in array inspection.

Q4.8: Section 2.1 and 2.2: Are all equations developed by the authors for the explanation of phased array and total focusing method? If not, cite them properly. --- The time of flight is T or TOF as used in Line 56.

A: In this paper the total focusing method is regarded as the limit case of conventional phased array method. On the other hand, the full matrix is regarded as the bases of vector space of linear ultrasonic methods. Based on these ideas, E.q. (1) to (4) are developed in my own languages.

Q4.9: Eq. 1-Eq. 4 are not cited. Are they developed by the authors themselves? If not, proper citations are required.

A: In this paper the total focusing method is regarded as the limit case of conventional phased array method. On the other hand, the full matrix is regarded as the bases of vector space of linear ultrasonic methods. Based on these ideas, E.q. (1) to (4) are developed in my own languages.

Revision of section 3 (Theoretical analysis)

Q5.1: Please correct throughout the text that the velocity of propagating waves (Shear /longitudinal) in Aluminium instead of the velocity of Aluminium.

A: Old sentences “The velocity of wedge is 2337m/s. The longitudinal and shear velocity of Aluminum are 6330m/s and 3080m/s.” has been revised to “The longitudinal wave velocity in wedge is 2337m/s. The longitudinal and shear wave velocity in Aluminum are 6330m/s and 3080m/s.” between line 138 to 139.

Q5.2: Please mention the SNR value of each Image if you are comparing the SNRs.

A: A new indicator named Defect SNR (DSNR) has been introduced in this paper. The definition is given in E.q. (7). All analyses of imaging performance of different modes are re-written using DSNR, mean of DSNR and the standard deviation of DSNR. They are given in Table. 3 to 6.

Q5.3: Please prepare a Table to compare the actual size of defects and estimated size with wedge coupling.

A: Table. 7 has been added to give the actual size of defects and estimated size with wedge coupling. Besides, the linear sizing equation for lower half region rectangular defect using linear regression are given at the same time.

Revision of section 4 (Experimental analysis)

Q6.1: Please add the photo of object and array of transducers in Figure 14.

A: Object and array of transducers has been added into Figure. 14.

Q6.2: Compare the experimental and numerical results in terms of defect size with accuracy in measurement.

A: According to the theoretical analysis result, the defect size has effect only for defect larger than 1mm. Unfortunately, the defect size is smaller than this minimum length (section 4.1). So it’s unnecessary to Compare the experimental and numerical results in terms of defect size with accuracy in measurement.

Revision of section 5 (Conclusion)

Q7.1: The conclusion section should be fully rewritten to show what the main input of this work to this area in a concise way.

A: The conclusion has been re-written.

Revision of Others

Q8: there are lots of variables and abbreviations in this manuscript, a Nomenclature Section should be added.

A. Table. 1 has been added in page 3 to introduce some variables and abbreviations in this manuscript.

Round  2

Reviewer 1 Report

Even if the authors have improved the paper, I have few remarks on this version of the paper :

R1 :

“the thicknesses of many thin-walled structure will be 10 times smaller than the minimum longitudinal wave-length of available array transducer.”

“The aim of this paper is to analyze the performance of linear ultrasonic array methods for thin-walled structure whose thickness is 5 times smaller than the longitudinal wave-length”

These statements are not coherent with the dimensions of the plate and the center frequency of the transducers… the thickness is 5 times bigger than the longitudinal wavelength (around 1.3 mm).

R2:

The figures show a scale between 0 and -25dB but the DSNR can go up to 50dB.

R3:

Table 7 seems to be for SS-S mode. It should be notified.

Author Response

To reviewer

Several typos appeared in manuscript have been revised, and the analyses of experimental results using DSNR have also been added in section 4.

R1: These statements are not coherent with the dimensions of the plate and the center frequency of the transducers… the thickness is 5 times bigger than the longitudinal wavelength (around 1.3 mm).

A: I had made several big mistakes in the descriptions of the relationship between thickness and wave-length. Sentences like “be 10 times smaller than” and “be 5 times smaller than” have all been replaced by “be smaller than 10 times of” and “be smaller than 5 times of” in the round2 revised manuscript.

R2: The figures show a scale between 0 and -25dB but the DSNR can go up to 50dB.

A: The image amplitude in this paper is rescale by the color scale before displaying. Suppose the color scale range is [0, -T]. For pixel with amplitude lower than -T, the final amplitude used for display will be set to -T which is the minimum threshold. So, the DSNR can go up to 50dB while figure shows a scale 0 and -25dB.

R3: Table 7 seems to be for SS-S mode. It should be notified.

A: The notification “SS-S mode” of Table. 7 has been added to the title.

Reviewer 2 Report

Comments for author:

Thank you for putting your best efforts and revising the manuscript. All comments have been addressed. There are following minor revisions which must be taken in account before considering to be published in the journal of “Coating”:

·         It is unnecessary to include figures, equation numbers etc. in Conclusion. You have to write only conclusive summary of article in short.

·         Throughout the text, please change as Fig. ( ) shown by “ as shown in Fig. ( ).

·         There are still mistakes in sentence structure of English and therefore it is recommended to take help of English teacher /native English speaker or English proofreading facility by publisher. This is very important to improve the quality of this article.

Author Response

To reviewer

Several typos appeared in manuscript have been revised, and the analyses of experimental results using DSNR have also been added in section 4.

R1: It is unnecessary to include figures, equation numbers etc. in Conclusion. You have to write only conclusive summary of article in short.

A: The conclusion has been revised again according to review’s advice.

R2: Throughout the text, please change as Fig. ( ) shown by “ as shown in Fig. ( )..

A: Sentences with pattern “as Fig. ( ) shown by” throughout the manuscript have all been replaced by pattern “as shown in Fig. ( )”.

R3: There are still mistakes in sentence structure of English

A: The sentences with inappropriate structure of English I can find have been revised.