Deflection Monitoring Method for Simply Supported Girder Bridges Using Strain Response under Traffic Loads
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsAn interesting work is presented in this paper, validated by both numerical and field experiments. Long-gauge Fiber Bragg Grating sensing is utilized in this paper to obtain bridge deflection from strain measurements. Overall, the paper is well-written and presented, but a few comments need to be addressed before acceptance.
A key question is, is it feasible to capture the bridge damage, like the crack damage, by using the proposed approach? with the given accuracy, i.e., around 5%, or even near 10%, relative errors.
In Section 4.1, it’s suggested to provide a table here, to include and present all the bridge properties, e.g., all the dimensions, stiffness, cross-section details, etc.
In Figure 13, the wavelet transform based denoising technique must be explained in detail. Also, different line types are suggested to present different results in Figure 13, and all other figures presented in this paper.
In Section 4.4, different parameters are investigated, including vehicle type, vehicle speed, and structural damage, how about the interaction between the bridge and the crossing vehicle, as well as the road roughness? The air, and surface temperatures? Some discussions are needed here, according to some recent works, i.e., DOI: 10.1111/mice.12974, 10.3390/buildings12081209, along with some FBG related works, e.g., 10.1109/JSEN.2014.2332874, 10.1109/JSEN.2013.2256599, are suggested to strengthen the introduction.
Sampling frequency for all the FBG is suggested to be provided in this paper, i.e., Sections 4, and 5.
In Section 3.3, a typo is noticed here, just at the beginning of the subsection, a bracket is needed for the reference [25].
Author Response
Comment 1. it feasible to capture the bridge damage, like the crack damage, by using the proposed approach? with the given accuracy, i.e., around 5%, or even near 10%, relative errors.
Answer: The proposed method is applied to assess the bridge deflection under the traffic loads as the deflection is usually selected as a key parameter to evaluate the global health of the bridge. The deflection itself is not sensitive to local damage, such as crack damage. If the local damage is needed to be identified, the method can be referenced as ‘Yongsheng Tang, Zhishen Wu, Caiqian Yang, Gang Wu. A model-free damage identification method for flexural structures using dynamic measurements from distributed long-gage macro-strain sensors. Journal of Intelligent Material Systems and Structures, 2014, 25(13): 1614-16’.
Comment 2. In Section 4.1, it’s suggested to provide a table here, to include and present all the bridge properties, e.g., all the dimensions, stiffness, cross-section details, etc.
Answer: It has been added in the submission as shown in Figure 10 and Table 1.
Comment 3. In Figure 13, the wavelet transform based denoising technique must be explained in detail. Also, different line types are suggested to present different results in Figure 13, and all other figures presented in this paper.
Answer: The detailed information of the denoise method has been added in the paper. The figures have also been improved.
Comment 4. In Section 4.4, different parameters are investigated, including vehicle type, vehicle speed, and structural damage, how about the interaction between the bridge and the crossing vehicle, as well as the road roughness? The air, and surface temperatures? Some discussions are needed here, according to some recent works, i.e., DOI: 10.1111/mice.12974, 10.3390/buildings12081209, along with some FBG related works, e.g., 10.1109/JSEN.2014.2332874, 10.1109/JSEN.2013.2256599, are suggested to strengthen the introduction.
Answer: Actually, there are many other parameters influencing the strain as the comment presents. The air and surface temperatures often show large influence on the long-term static strain measurements. Luckily, the strain measurement here will be implemented in very short time under the traffic loads. The road roughness will make some additional vibration, which can be denoised with wavelet transform. For the crossing vehicle, it is really affecting the proposed method. Therefore, the proposed method is mainly applied to calculate the deflection under the case of traffic loading on only one lane. The case can be identified with the strain sensors installed at each lane.
Comment 5. Sampling frequency for all the FBG is suggested to be provided in this paper, i.e., Sections 4, and 5?
Answer: The sampling frequency is for all the FBG 200 Hz. It has been modified in the article.
Comment 6. In Section 3.3, a typo is noticed here, just at the beginning of the subsection, a bracket is needed for the reference [25].
Answer: It has been modified.
Reviewer 2 Report
Comments and Suggestions for AuthorsComment 1
- The abstract could benefit from a more concise and clearer presentation of the proposed method. Some sentences are complex, and the overall structure might be refined for better readability.
Comment 2
- Authors mention an "approximately 5%" calculation error, which might be perceived as relatively high in engineering assessments. More context on industry standards or comparisons with existing methods could provide a clearer perspective on the proposed accuracy.
Comment 3
- The introduction does not sufficiently discuss or compare the proposed method with existing deflection measurement techniques, such as geodetic methods, LVDT, GPS, or acceleration sensors. A more comprehensive evaluation of the proposed method against these alternatives would strengthen the justification for its novelty and effectiveness.
Comment 4
- The paper briefly mentions limitations of existing methods but fails to delve into potential drawbacks or challenges associated with the proposed method. Addressing possible limitations, such as environmental conditions affecting strain measurements or specific scenarios where the method might not be suitable.
Comment 5
- Paragraph (4.2) lacks a clear justification for choosing 20 measurement units, each 1 m long. It's crucial to provide a robust rationale for this specific division, considering factors such as the bridge's structural characteristics, expected strain variations, or industry standards. Without such justification, the chosen configuration appears arbitrary and may raise questions about the method's adaptability to different bridge types or conditions.
Comment 6
- The use of the term "imagined sensor" is unclear and introduces ambiguity.
Comment 7
- For the employed wavelet transform to denoise the strain signal, in order to enhance the transparency and reproducibility of the study, it would be necessary to furnish explicit details about the configuration of the wavelet transform, such as the chosen wavelet function, the number of decomposition levels, and the thresholding criteria applied.
Comments for author File: 
Comments.pdf
Author Response
Comment 1. The abstract could benefit from a more concise and clearer presentation of the proposed method. Some sentences are complex, and the overall structure might be refined for better readability.
Answer: It has been modified.
Comment 2. Authors mention an "approximately 5%" calculation error, which might be perceived as relatively high in engineering assessments. More context on industry standards or comparisons with existing methods could provide a clearer perspective on the proposed accuracy.
Answer: For civil engineering, an "approximately 5%" calculation error is accepted as the deflection value is not large, often smaller than several mm. Therefore, the deflection measurement error is about 0.05 mm or 0.1 mm. It is still very meaningful especially for the short- or medium-span bridges as the long-term monitoring cost is often very small.
Comment 3 The introduction does not sufficiently discuss or compare the proposed method with existing deflection measurement techniques, such as geodetic methods, LVDT, GPS, or acceleration sensors. A more comprehensive evaluation of the proposed method against these alternatives would strengthen the justification for its novelty and effectiveness.
Answer: It has been modified.
Comment 4. The paper briefly mentions limitations of existing methods but fails to delve into potential drawbacks or challenges associated with the proposed method. Addressing possible limitations, such as environmental conditions affecting strain measurements or specific scenarios where the method might not be suitable.
Answer: It has been added ‘Besides the sensors, another factor, namely strain–deflection relationship, will determine the performance of these indirect method. The strain–deflection relationship is usually determined by the bridge structure mode. If the difference between the mode and actual structure is small, the deflection assessment accuracy can be promised. If the difference is large, the method may fail. Finally, the factors, such as temperature, loading and damage, causing strain change, may also influence the performance of the indirect method based on strain input. Therefore, the method should be further studied for actual application.’
Comment 5. Paragraph (4.2) lacks a clear justification for choosing 20 measurement units, each 1 m long. It's crucial to provide a robust rationale for this specific division, considering factors such as the bridge's structural characteristics, expected strain variations, or industry standards. Without such justification, the chosen configuration appears arbitrary and may raise questions about the method's adaptability to different bridge types or conditions.
Answer: The former researches results show that the gauge length of the sensor should not larger than 1/20 of the girder length to obtain an accurate displacement measurement when the sensors are distributed install at the girder. The detail can be found in reference ‘Tao Wang, Yongsheng Tang. Dynamic displacement monitoring of flexural structures with distributed long-gage macro-strain sensors. Advances in Mechanical Engineering, 2017, 9(3): 1-12’
Comment 6. The use of the term "imagined sensor" is unclear and introduces ambiguity.
Answer: It has been modified ‘Here, there is no actual sensor model in FE model, while the strain is just extracted at the position as the designed sensor lies. For each sensor, the average strain within the gauge length is obtained from the model directly, which is taken as the measured macro-strain.’
Comment 7. For the employed wavelet transform to denoise the strain signal, in order to enhance the transparency and reproducibility of the study, it would be necessary to furnish explicit details about the configuration of the wavelet transform, such as the chosen wavelet function, the number of decomposition levels, and the thresholding criteria applied.
Answer: It has been modified.
Round 2
Reviewer 1 Report
Comments and Suggestions for Authors
All comments have been addressed by the authors. The manuscript can be accepted in its current form.
Reviewer 2 Report
Comments and Suggestions for AuthorsThe authors have revised the manuscript as per my comments
