Damage Detection in Reinforced Concrete Member Using Local Time-Frequency Transform Applied to Vibration Measurements
Round 1
Reviewer 1 Report
This paper applied the local time-frequency transform (LTFT) to calculate the time-domain signals. Both simulated as well as signals collected from a large-scale test of a reinforced concrete girder were processed with the LTFT and compared with Renyi entropy to quantify the time-frequency spectrum, the time-frequency resolution abilities of short time Fourier transform (STFT), and S transform(ST). The results show that this method has high-precision time-frequency resolution and excellent noise robustness. Although the article falls within the scope of the journal, I do not recommend publishing this paper in its current form. The specific comments are shown as follows:
1. The authors need to carefully improve the writing of this paper. The English of this paper is very hard to read. For instance, consider the abstract, “The results show that the LTFT is superior to the traditional time-frequency analysis schemes, in that it is more effective in identifying the energy changes in the time-frequency spectrum before and after structural damage in form of cracking has occurred. At the same time, it provides high-precision time-frequency resolution and excellent noise suppression. The effectiveness and feasibility of the LTFT applied to the synthetic and experimental signals are verified.” The Reviewer does not understand what the Authors are trying to say. What does “noise suppression” even mean? There are many such written statements in the paper. The readability of this paper needs to be improved significantly before its final publication.
2. Revise the introduction section. Especially the first statement of the abstract, “Many signal processing techniques such as time-domain, frequency-domain and time-frequency domain analysis have been used in vibration-based SDD and SHM research. Some of the most common methods of time-frequency domain analyses are briefly reviewed in this section.” These two sentences seem to have no logical relationship, which is very abrupt. Please review the whole paper and revise it.
3. The whole keywords needs to be rewritten. The content of this research should be clearly presented in the keywords from specific to extensive.
4. In this paper, the effectiveness and feasibility of the LTFT applied to the synthetic and experimental signals are verified. Author mentioned that compared with short time Fourier transform (STFT), and S transform(ST), the method has high-precision time-frequency resolution and excellent noise suppression. However, there is no quantitative comparison of results with significant findings, which may lack of persuasion.
5. There are many loosely written statements in the paper, for example, in subsection 2.1, author uses “Letting” many times, and in line 521-522, “Figure 14 shows the stacked spectra before and after damage at each location, for comparison.” The Authors need to revise the whole paper in terms of writing significantly to avoid rejection.
6. Author used Figure 14 to show the stacked spectra before and after damage at each location. , but its content is the same as that in Figure 12. Please consider reorganizing this content to make it clear.
7. Rewrite the conclusion section. In the Conclusion section, the conclusions abstracted by the authors are not refined, the authors should re-write the conclusion section and illustrate the main conclusions and quantitative comparison of the results of this paper. Additionally, the authors should include the main limitations of the work presented in the paper.
Author Response
Please see the attachment
Author Response File: 
Author Response.doc
Reviewer 2 Report
This manuscript describes a time-frequency analysis method applied to vibration signals for structural damage detection, which is a highly concerned topic in civil infrastructures. However, there are some issues which need to be addressed before it’s accepted.
1) The time-frequency spectrum appeared many times in the manuscript, but the physical meaning expressed by different color levels was not introduced in detail. Please add the corresponding description.
2) Several different time-frequency analysis methods were introduced, and the relationship between these methods should be described in more detail.
3) There is typo in line 457, highs-speed.
4) Line 458, a sampling frequency of 100 kHz is not reasonable for vibration tests. According to Figure 14, the sampling frequency seems to be 5 kHz. Please check this value.
5) In table 2, the relative changes of frequencies would be more meaning.
6) Comparing with the direct frequency changing, the Rényi entropy is more sensitive for damage detection?
7) Please explain why the Rényi entropy of STFT does not decrease after structural damage.
Author Response
Please see the attachment.
Author Response File: Author Response.doc
Reviewer 3 Report
The Authors focused on signal processing and analysis of structural vibration measurements based on structural damage detection in structural health monitoring. The Authors’ goal was to extract subtle changes in the measured signals, which can be used to infer changes in structural parameters and damage. The article seems to be interesting for a wide array of scholars and engineers. It creates logical scientific research and that is why in my opinion could be published in "Buildings" after minor revision. Some of the comments on the manuscript are listed below.
1) Figure 1÷3; the unit of amplitude along the ordinate is missing. Please, add the units.
2) Equation (2); why the Authors claim to name the equation (2) the complex exponential form? The Reviewer cannot see any exponents in the mentioned equation.
3) Equation (4); why does the summation index k start from zero? why b0 = 0? Perhaps, the Authors intended to start with index k=1 and with a0 = 0 (instead of b0 = 0), so in this case the equation (4) could correspond to the earlier given equation (1).
4) Equation (5); the Reviewer agrees to some degree with the Authors that coefficients Ck represents the dot product, however the Authors did not mention the integration process. From the Reviewer point of view the dot product without integration will not work and will not allow the Authors to obtain the correct coefficients Ck. Please, explain why the Authors avoided integration process during finding these coefficients Ck.
5) Equation (8); perhaps, the time function should be numbered too.
6) Line 298; why the frequency changes until 120 Hz? The initial value of 20 Hz is understood but the upper utmost limit is a puzzle for the Reviewer. How do the Authors could explain that upper limit equals 120 Hz?
7) The equations (1÷22); if the equations are taken from literature, then the citations are required (some equations are cited by Authors but some of them are not).
8) Figure 1÷8 and 10÷16; the numbers as well as the label descriptions along the abscissas and ordinates should have larger font size. The size should be comparable with the main text of the manuscript. Please enlarge it because in the current form is too small.
Author Response
Please see the attachment.
Author Response File: Author Response.doc
