Vibration-Based and Near Real-Time Seismic Damage Assessment Adaptive to Building Knowledge Level

Round 1

Reviewer 1 Report

The authors review many damage identification methods, including empirical, analytical and numerical methods. They study the accuracy of these methods using an experiment, and propose a framework to assess the seismic damage grades. The study is significant for future real-time damage assessment.

Some questions are raised below:

1.  The authors evaluate the accuracy of various methods by comparing the results to the observation (listed in Table 1 and described in 2.2). But, how accurate is the reference (observation)?
2. Section 3.3, Line 483: The numerical method could not provide accurate results until non-linear part. The results are very interesting, but how to explain this?
3.  Line 540: According to the authors, the ease and speed of the method also matter for rapid damage assessment. Can the authors quantify the computational time and compare them in Table 4 as another parameter?
4. Many format errors present, for example, number of sections and figures.

Author Response

Dear reviewer,

Many thanks for your valuable time and consideration. Please kindly refer to the attached response letter.

Author Response File: Author Response.pdf

Reviewer 2 Report

The paper presents an interesting and quite relevant approach for fast seismic damage assessment and does a step forward contribution to this topic.

However, some aspects should be improved and dealt with in a little more detail.

First, authors should a bit elaborate the concept of knowledge level (source, idea, references). The difference between KL1 and KL2 does not become very clear. Especially with respect to the amount of required sensors / used sensors; instrumentation schemes and the real-world application (use of earthquake response measurements). So far, the paper and the results rely only on one controlled shake table experiment!

Figure 1 is a bit misleading. The link / path between the KL and methods does not become clear.

The empirical method (figure 1a) shows a quite broad scatter. Whereas Figure 1b seems illogic. If the ratio between fi / f0 = 1: No Damage occurred! Whereas Figure 1b indicate probability of DG 4 = 25% ??

In case of the presented Method 2, the authors need to explain the derivation of the initial bilinear curve for the case study building. Especially in the context of the KL!

Readers cannot understand the derivation of the results presented in Table 4 because information is missing.

Figure 6: The color scale is missing

A table at the end of section 1.1 summarizing the differences between the methods (e.g. inputs, outputs) would help the reader better understand the utility of using these different methods.

Line 100: The distinguish between KL2, and KL3 could be better expressed in a tabular form where specific information is specified to help distinguish the two levels (e.g. Number of floors, layout, construction material properties).

The list of references needs to be checked with the references mentioned in the text (e.g. Line 59, Rytter (1993), Line 244 Priestley et al. (2007) are missing in the list of references).

Consistency with abbreviations use is needed (e.g. EFDD in Line 110 and (E)FDD in line 496).

Figures numbering need to be revised (e.g. Line 471, 493, 503) and the style of the figure title need to be uniform (bold name, and the period in the end.

Chapter numbering need to be checked

Line 365 Separate the table and figure titles.

Figure 4(b), line 471, it’s not clear how test6, test8, and test9 are related to the signals above them, and the x-axis label needs to be changed to (time(s)) to be uniform with the other figures.

line 508, use s instead of sec as time unit.

Author Response

Dear Reviewer,

Thank you for your kind consideration of our manuscript. Please see the attached file for our revisions addressing your comments.

 

Author Response File: Author Response.pdf

Reviewer 3 Report

The authors proposed a multi-level framework for near real-time earthquake damage assessment. In the article, some dynamic identification methods and damage assessment techniques are shown. In addition, an open source data from a multi-story building structure was used to validate the proposed framework. The overall article is very well organized. However, I still have few doubts before publishing this article.

1. The authors claim that this is a near real-time framework. How this near real time is defined, it is suggested to give a specific time for earthquake damage assessment.
2. The authors demonstrate three levels of building knowledge in their paper. However, in actual engineering practice, KL1 is one of the most practical approaches. Although the framework proposed by the authors is very complete, the importance of KL1 should be highlighted.

Author Response

Dear Reviewer,

Many thanks for your thorough comments in helping us improve the quality of the manuscript. Please see the attached file to view the updates we have made relevant to your queries.

 

Author Response File: Author Response.pdf