Analysis of the Recorded Response of a School Building Heavily Damaged by the 2016 Central Italy Earthquake
Round 1
Reviewer 1 Report
The paper addresses an interesting topic, the capability of combined use of experimental response and computational results in order to quantifying damage in structures.
The paper is well written and well organized. Nevertheless, some clarifications are required to improve the quality of the paper before publication.
The introduction is clear and introduces into perspective what follows. But the authors should summarize recent research related to the combination of experimental and numerical results for the determination of structural damage. And it is important to highlight the novelty and topicality of the manuscript.
The authors indicate “A 20 minutes ambient vibration measurement recorded in 2010 and a 1 hour ambient vibration measurement before the 2016 earthquake are available.” (line 165). How can they prove with only two measurements of such short duration that the variations are due to damage?
The authors do not describe the software used in numerical analysis or the characteristics of the numerical model.
They explain that the roof trusses have been modelled using frame structures but figure 6 does not depict this. (line 200)
What method was used to update the model and minimize the error? (line 221)
On the basis of which analysis do they indicate that a 10% increase in the elastic modulus is reasonable? (line 248)
The authors write “In any case, the result obtained was considered sufficiently approximate to use this model for the simulations discussed in the following Section.” (line 258). Why is sufficiently approximate?
The conclusions section summarizes the document and describes the results. The conclusions should reflect upon the aims. If no general conclusions are defined and only specified on the example, the article is a case study.
Please, take care to indicate the name of the acronym the first time it appears in the text (line 78, line 119…)
Please, check reference 7.
Author Response
Question --> The paper addresses an interesting topic, the capability of combined use of experimental response and computational results in order to quantifying damage in structures. The paper is well written and well organized. Nevertheless, some clarifications are required to improve the quality of the paper before publication.
Answer --> The required clarifications are reported in the following.
Q --> The introduction is clear and introduces into perspective what follows. But the authors should summarize recent research related to the combination of experimental and numerical results for the determination of structural damage. And it is important to highlight the novelty and topicality of the manuscript.
A --> The following part has been added (line 70): "As well known, structural health monitoring (SHM) techniques can be applied on data gathered from a seismic monitoring systems. Even if a SHM comprehensive review is outside the aim of this paper, the classification reported in (Gharehbaghi, 2022) is of interest, where the approaches proposed in the literature are distinguished in model-based or signal-based. In the former approaches, damage is recognized through tracking variations in the simulated measurements from a structural model and specific parameters of this model are updated under the system's responses, as discussed for example in (Benedettini, 2017). The latter approaches rely on statistical analysis and assess the system's response independently; therefore, they do not require additional information concerning the structure's physical properties and parameters. One of the main issues of signal-based techniques is the limitation of the number of response quantities. As discussed in the following, combining experimental and numerical results, available response quantities can be expanded in order to obtain much information at local level, thus allowing the estimation of damage on all relevant structural elements."
At the end of the Introduction the following phrase has been added: "This approach is useful also in a SHM perspective, because it allows to expand the experimental data set and to gather information at local level on a great number of structural elements. "
Q --> The authors indicate “A 20-minute ambient vibration measurement recorded in 2010 and a 1-hour ambient vibration measurement before the 2016 earthquake are available.” (line 165). How can they prove with only two measurements of such short duration that the variations are due to damage?
A --> In this part of the paper, the ambient vibration measurements are used to estimate the changes in the linear range, owing to the modifications due to the structural upgrading interventions performed in 2012. The signal/noise ratios of these measurements were analyzed and they resulted adequate to allow a reliable estimation of the building's dynamical properties and of their corresponding changes. The following phrase has been added at the end of the paragraph: "These data sets will be denoted respectively by 2010 and 2016 and can be suitably analyzed using linear procedures, in order to determine the updated model after the 2012 interventions."
Q ---> The authors do not describe the software used in numerical analysis or the characteristics of the numerical model.
A ---> The citation of the commercial software has been added.
Q ---> They explain that the roof trusses have been modelled using frame structures but figure 6 does not depict this. (line 200)
A ---> The following phrase has been added: "the roof trusses have been modelled using frame elements, while the corresponding plane elements in Fig. 6a were used only to simulate the masses."
Q ---> What method was used to update the model and minimize the error? (line 221)
A ---> The following phrase has been added: "In the framework of structural identification procedures, model updating techniques usually involves suitable analytical nonlinear procedures to drive the optimum selection of the parameters (De Sortis, 2005, 2007). When the focus of the study, like in the present case, is on the analysis of the seismic performance, rather than on the analytical features of the method, a linearized modified procedure based on (Douglas, 1982) can be effectively applied."
Q ---> On the basis of which analysis do they indicate that a 10% increase in the elastic modulus is reasonable? (line 248)
A ---> Recalling the previous observation that the proposed paper concerns the analysis of the seismic performance, rather than the analytical features of the identification method, in this case an analytical model updating technique was not applied. Instead, the 2010 model was modified to reflect the 2012 interventions. This part has been modified as follows: "As already discussed, in 2012 some structural interventions were made. Thus, the FEM2010 model has been modified, in order to approximate the new dynamical parameters reported in Tab. 1 under the column "Experimental 2016". The new model, called FEM2016, reflects those structural modifications: deletion of the connections between buildings A and C at all floors, except at the roof, analogously to building B in the FEM2010 model, and a 10% increment to the elastic modulus of the masonry infills, due to the reinforced plaster applied to the walls."
The caption of Table 1 has been accordingly modified: "FEM 2016: modification of FEM 2010 to reproduce the effects of 2012 interventions".
Q ---> The authors write “In any case, the result obtained was considered sufficiently approximate to use this model for the simulations discussed in the following Section.” (line 258). Why is sufficiently approximate?
A ---> The 2010 model was not suitable to simulate the 2016 seismic response, because in 2012 the structural configuration was changed. Thus, a model updating has been pursued leading to the 2016 model. Notwithstanding the 2016 model shows greater errors with respect to 2010, however it is in good agreement with the new experimental data, thus it was judged sufficiently approximate to be used in the prediction of the response.
The phrase has been slightly modified: "In any case, the result obtained was considered sufficiently approximate to use this model for the kind of simulations discussed in the following Section."
Q ---> The conclusions section summarizes the document and describes the results. The conclusions should reflect upon the aims. If no general conclusions are defined and only specified on the example, the article is a case study.
A --->The following phrase has been added: "In a SHM perspective, in this paper the use of experimental results and nonlinear models has been combined: in particular the numerical model has been used, along with the experimental results, to obtain information at local level, thus allowing the estimation of damage on all relevant structural elements. This technique can be very useful in general to obtain more significant results from the recorded response.”
Q ---> Please, take care to indicate the name of the acronym the first time it appears in the text (line 78, line 119...)
A ---> All abbreviations have been checked and corrected.
Q ---> Please, check reference 7.
A ---> The reference has been checked and corrected.
Reviewer 2 Report
This paper described a finite element analysis and field study of a school building under seismic loads. The elastic behaviour of the building was validated by dynamic properties obtained from field measurement. The inelastic response under destructive earthquake was calculated from finite element analysis. However, the computer model is unclear. My comments on the non-linear model are listed below.
1) The main building considered was made of a masonry including irregular stones and clay bricks. The floor slabs consisted of hollow clay with RC joists. The adjacent buildings were constructed with reinforced concrete. The authors should clearly present the non-linear material properties of key construction materials considered and the nonlinear models of various structural components adopted in the analysis.
2. The RC and connection details have not been presented. The numerical simulation of those details should be given.
3. Failure model of the building in the non-linear analysis should be presented and compared with the post-earthquake observation to validate the computer model.
The technical contents of this paper should be substantially improved before it could be accepted for publication.
Author Response
Question ---> This paper described a finite element analysis and field study of a school building under seismic loads. The elastic behaviour of the building was validated by dynamic properties obtained from field measurement. The inelastic response under destructive earthquake was calculated from finite element analysis. However, the computer model is unclear. My comments on the non-linear model are listed below. The technical contents of this paper should be substantially improved before it could be accepted for publication.
Answer --> The required clarifications are reported in the following
Q ---> 1) The main building considered was made of a masonry including irregular stones and clay bricks. The floor slabs consisted of hollow clay with RC joists. The adjacent buildings were constructed with reinforced concrete. The authors should clearly present the non-linear material properties of key construction materials considered and the nonlinear models of various structural components adopted in the analysis.
A ---> The nonlinear analysis is restricted, in this paper, to RC building A. The following phrase has been added: "Material strength values have been obtained during the already reported survey, performed after the installation of the monitoring system."
Q ---> 2) The RC and connection details have not been presented. The numerical simulation of those details should be given.
A ---> This phrase has been added to clarify: "To this aim, the recorded displacements were applied to the points of the model corresponding to the location of the sensors; the plane stiffness of the floor slabs allowed to transfer these displacements to the remaining points. It can be easily recognized that this procedure does not require the knowledge of the nonlinear interaction between buildings C and B."
Q ---> 3) Failure model of the building in the non-linear analysis should be presented and compared with the post-earthquake observation to validate the computer model.
A ---> The following phrase has been added: "This nonlinear model does not have a predictive aim, in order to simulate the failure behaviour, but its aim is the use of measured response for the estimation of local damage quantities."
Reviewer 3 Report
The paper presents a study on the monitored response of an existing school buildings damaged during the Amatrice earthquake. The paper is interesting and fits with the scope of the Journal. In this Reviewer's opinion, the paper shall be considered for publication after that some aspects will be clarified. Please, see the comments in the attached PDF file
Comments for author File: 
Comments.pdf
Author Response
Question ---> Line 58: I suggest to do not indicate in this way a group of papers, but authors could mention the first names of the authors of the references.
Answer ---> The citation style follows that of the Journal. In this part of the paper, 19 studies are cited, thus to cite the names of all authors seems too space-consuming.
Q ---> Line 82: I suggest to enclose a brief state of the art about the importance of school buildings, as one of categories in the existing building stock that deserve attention. Following some suggestions of the main recent literature about the topic, which authors could take into account:
https://doi.org/10.1016/j.ijdrr.2022.102924
https://doi.org/10.1007/s10518-020-01033-5
https://doi.org/10.5194/nhess-19-1365-2019
https://doi.org/10.1007/s10518-020-00989-8
https://doi.org/10.1016/j.engstruct.2018.04.056
https://doi.org/10.1016/j.ijdrr.2020.101807
A ---> The following phrase has been modified at the beginning of Section 2, adding some citations: "The school building aggregate studied in this paper, belonging to a typology widely studied in a seismic perspective in the last years (Vatteri, 2022, Ruggieri 2020), is located near the centre of Amatrice."
Q ---> Line 90: maybe, it is better indicate the slab as ribbed.
A ---> The phrase has been modified as follows: "The original wooden roof was substituted in the 70s by ribbed slabs, similar to those of the floors."
Q ---> Line 98: To be honest, this classification is too old (it dates back to 2001, even if it is actually used in the practice). Perhaps, you could provide an equivalence of the seismic zone with a PGA range in order to provide a better indication for not Italian readers. In addition, if you account for zones, also zone 4 should be mentioned
A ---> This phrase has been modified as follows: "In Italy three seismic zones exist, according both to historical and modern regulations (Zone 1 corresponds to locations with the highest seismicity), not considering Zone 4 where the seismic hazard is negligible.
Q ---> Line 119: To define
A ---> The abbreviation has been defined in the text.
Q ---> Line 159: Major details on the monitoring systems are required. In addition, you should justify the choice of the monitored points within the floors.
A ---> The following phrase has been added: "In particular, the sensors locations were selected in order to allow the definition of three different rigid body constraints, as discussed in the following."
Q ---> Line 193: From the FEM model figure, it seems that the structural joints is not accounted within the model. Is this assumption correct in the numerical model?
A ---> The model was checked, this is only a visualization problem. Thus, the following phrase has been added: "due to visualization approximations, buildings C and B seem connected at all floors, but in the model they are connected only at the roof level".
Q ---> Line 199: Please, mention the software used.
A ---> The citation of the commercial software has been added.
Q ---> Line 219: How did you select the mechanical properties of the materials accounted in the numerical model?
A ---> The following phrase has been modified: "The a priori model characteristics were based on the geometry and material properties, purposely investigated after the installation of the monitoring system".
Q ---> Line 322: This sentence is not really clear. Please, add major details
A ---> The phrase has been clarified as follows: "As expected [12–21,33], FEM2016 model is able to reproduce the real behaviour until the elastic limit is exceeded, after that a significant period elongation of the structure is evident in the experimental results. Nonetheless, the elastic model is capable of predicting rather well maximum accelerations at various locations (Tab. 4)."
Q ---> Line 354: Also in this case, major details shall be added.
A ---> The following phrase clarifies this point: "Nonlinear models of RC buildings have been developed, according to the lumped plasticity approach for the monodimensional elements, beams and columns, using the "hinge properties" definition of the code [26]." Moreover, in the subsequent part of the text more details were already reported.
Q ---> Line 395: Honestly, it is not really clear what is the aim of the nonlinear model proposed in the end of the paper. What did authors want to demostrate?
A ---> The following phrase has been added: "In a SHM perspective, in this paper the use of experimental results and nonlinear models has been combined: in particular the numerical model has been able, making use of the measured displacements, to give information at local level, thus allowing the estimation of damage on all relevant structural elements."
Round 2
Reviewer 2 Report
My previous comments have been addressed.
