Behavior of Barrier Wall under Hydrogen Storage Tank Explosion with Simulation and TNT Equivalent Weight Method
Round 1
Reviewer 1 Report
Major revision has to be performed in order to make the paper easier to read and to clarify certain points.
Main points:
1) Explain why for this study a 75% ratio of hydrogen with air was taken.
2) Why do you take explosive energy of TNT = 4.6 MJ/kg instead of generally accepted 4.2 MJ/kg?
3) Could you please give and explanation as to why the authors of [27] consider a value of 0.01 for the explosion efficiency (Eq. 1). Was it based on experimental data or numerical analysis?
4) For the LS-DYNA material models it would be better to provide corresponding formulae for the MAT_JHONSON_COOK and other models. For a reader who is not familiar with these models, the Table 4 and 5 are not informative. Same for the HIGH_EXPLOSIVE_BURN – need to present the analytical version of the model.
5) What is the justification for the choice of the mesh sizes (100 mm in air and 20 mm inside the structure). Have you performed grid sensitivity studies? If Yes, please provide the results.
Minor remarks:
a) The phrase (lines 11-12) is not clear. Please revise.
b) Line 43-44: the lower and upper limits for combustion (not laminar flame).
c) Line 194: hydrogem..
d) Line 211: In figure 2 i_s is….
e) Line 220: “The model was modelled…” Please revise.
Line 315: …table 7
Author Response
Response to Reviewer 1 Comments
Thank you for review
Point 1
Explain why for this study a 75% ratio of hydrogen with air was taken
Response 1
Thank you for your comment.
In present of 4 % to 75 % ratio of hydrogen with air, explode can occur. If we take less than 75 %, the equivalent weight of TNT will be less. To create worse case scenario 75% was selected.
Point 2
Why do you take explosive energy of TNT = 4.6 MJ/kg instead of generally accepted 4.2 MJ/kg?
Response 2
Thank you for your comment.
Explosive energy was taken from this article. This article calculate the explosion energy of TNT.
Kwon, S.; Park, J. A Review of TNT Equivalent Method for Evaluating Explosion Energy Due to Gas Explosion.
Point 3
Could you please give and explanation as to why the authors of [27] consider a value of 0.01 for the explosion efficiency (Eq. 1). Was it based on experimental data or numerical analysis?
Response 3
Thank you for your comments.
It has been updated in manuscript.
Point 4
For the LS-DYNA material models it would be better to provide corresponding formulae for the MAT_JHONSON_COOK and other models. For a reader who is not familiar with these models, the Table 4 and 5 are not informative. Same for the HIGH_EXPLOSIVE_BURN – need to present the analytical version of the model.
Response 4
Thank you for your comment.
References has been added.
Point 5
What is the justification for the choice of the mesh sizes (100 mm in air and 20 mm inside the structure). Have you performed grid sensitivity studies? If Yes, please provide the results.
Response 5
Thank you for your comment.
Grid sensitivity analysis had not been studied but the mesh size was taken from the verification of previous studies.
Point 6
Minor remarks:
a) The phrase (lines 11-12) is not clear. Please revise.
Response 6 a
Thank you for your comment.
It has been corrected.
b) Line 43-44: the lower and upper limits for combustion (not laminar flame).
Response 6 b
Thank you for your comment.
It has been corrected.
c) Line 194: hydrogem..
Response 6 c
Thank you for your comment.
It has been corrected.
d) Line 211: In figure 2 i_s is….
Response 6 d
Thank you for your comment.
It has been corrected.
e) Line 220: “The model was modelled…” Please revise.
Response 6 e
Thank you for your comment.
It has been corrected
f) Line 315: …table 7
Response 6 f
Thank you for your comment.
It has been corrected.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
This manuscript numerically investigated the dynamic response of H-beam walls under blast loading. A total of 36 numerical models were developed to examine the effects of structural layout, wall thickness, blast distance and TNT mass factors. As the ALE method is very expensive and the numerical models in the manuscript had large dimensions, a very large and costly computational effort was performed by the authors. The manuscript is of high quality and the research can contribute to the study of the impact dynamics of RC structures, within the scope of Applied Sciences, and the reviewer recommends publication of the manuscript after responding the following comments.
- L76, Nassr et al., 2012, citation error.
- In the introductory section, a literature review on the numerical simulation of explosions in RC structures was missing.
- Eq.(1), L296, “X” should be “×”.
- In section 2.2, an original reference to the Friedlander equation should be added to the presentation of the overpressure relation for free air explosion waves for the convenience of the reader. e.g., (10.1098/rspa.1946.0046)
- L217, the names of a,b,c in figure 4 were duplicated.
- It would be better to present more details about the finite element model, such as the mesh discretization, the cross section of the RC wall for different concrete thickness, etc.
- L249, The KCC and CSC models were applied to simulate concrete, but an introduction to the models, or references to them, were missing. References to material constitutive models should to be added. For the KCC model: (10.1061/(ASCE)EM.1943-7889.0000952), (10.1016/j.cemconcomp.2022.104911); For the CSC model: (FHWA-HRT-05-062), (10.1016/j.compstruct.2022.116424).
- L255, the original work of Malvar should be cited. (10.1016/s0734-743x(97)00023-7)
- How was the fluid-solid coupling in the manuscript achieved? Was it by using the traditional *CONSTRAINED_LAGRANGE_IN_SOLID, or did it use *ALE_STRUCTURED_FSI in the structured ALE approach. please clarify.
- The reviewer did not find a description of the contact between the reinforcement, H-beam and concrete, please check.
- It is necessary to add the error (%) between the numerical analysis and the experimental results in Table 6.
- The ALE method of calculating blast wave propagation requires a very fine mesh, and the underestimation of peak pressure at 1600 mm might be the reason for this. The "KPa" in Table 7 should be "kPa"
- It is advisable to revise Figure 8 to make the graph easier to understand and the axis ranges need to be modified to fit the data.
- From Figure 10, it seems that the element erosion method was used, but the reviewer did not find a description in the manuscript, please explain.
- If the element erosion method was used, then the volume of the failed element can be calculated in LS-prepost, which would provide a quantitative analysis for this study.
- The effective plastic strain in Figure 10 was essentially an index of damage, with strain hardening between 0 and 1 and strain softening between 1 and 2. Please explain this historical variable in the revision to avoid misinterpretation.
- Scaled distances for different load cases should be calculated and provided.
- Is it possible to obtain advice on which structural forms and dimensions are safe at certain blast distances under the most adverse blast loading scenario?
- The reinforcement arrangement in the RC wall needs to be supplemented.
Author Response
Response to Reviewer 2 Comments
This manuscript numerically investigated the dynamic response of H-beam walls under blast loading. A total of 36 numerical models were developed to examine the effects of structural layout, wall thickness, blast distance and TNT mass factors. As the ALE method is very expensive and the numerical models in the manuscript had large dimensions, a very large and costly computational effort was performed by the authors. The manuscript is of high quality and the research can contribute to the study of the impact dynamics of RC structures, within the scope of Applied Sciences, and the reviewer recommends publication of the manuscript after responding the following comments.
Thank you for review
Point 1
L76, Nassr et al., 2012, citation error.
Response 1
Thank you for your comment.
It has been corrected.
Point 2
In the introductory section, a literature review on the numerical simulation of explosions in RC structures was missing.
Response 2
Thank you for your comment.
Literature section has been updated.
Point 3
Eq.(1), L296, “X” should be “×”.
Response 3
Thank you for your comments.
It has been corrected.
Point 4
In section 2.2, an original reference to the Friedlander equation should be added to the presentation of the overpressure relation for free air explosion waves for the convenience of the reader. e.g., (10.1098/rspa.1946.0046)
Response 4
Thank you for your comment.
It has been corrected.
Point 5
L217, the names of a,b,c in figure 4 were duplicated.
Response 5
Thank you for your comment.
It has been corrected.
Titile of the article had been corrected.
Point 6
It would be better to present more details about the finite element model, such as the mesh discretization, the cross section of the RC wall for different concrete thickness, etc
Response 6
Thank you for your comment.
It has been added and mesh discretization of the simulation had been missing for this study.
Point 7
L249, The KCC and CSC models were applied to simulate concrete, but an introduction to the models, or references to them, were missing. References to material constitutive models should to be added. For the KCC model: (10.1061/(ASCE)EM.1943-7889.0000952), (10.1016/j.cemconcomp.2022.104911); For the CSC model: (FHWA-HRT-05-062), (10.1016/j.compstruct.2022.116424).
Response 7
Thank you for your comment.
It has been corrected
Point 8
L255, the original work of Malvar should be cited. (10.1016/s0734-743x(97)00023-7)
Response 8
Thank you for your comment.
It has been corrected
Point 9
How was the fluid-solid coupling in the manuscript achieved? Was it by using the traditional *CONSTRAINED_LAGRANGE_IN_SOLID, or did it use *ALE_STRUCTURED_FSI in the structured ALE approach. please clarify.
Response 9
Thank you for your comment.
It has been updated
Point 10
The reviewer did not find a description of the contact between the reinforcement, H-beam and concrete, please check.
Response 10
Thank you for your comment.
It has been updated
Point 11
It is necessary to add the error (%) between the numerical analysis and the experimental results in Table 6.
Response 11
Thank you for your comment.
It has been updated
Point 12
The ALE method of calculating blast wave propagation requires a very fine mesh, and the underestimation of peak pressure at 1600 mm might be the reason for this. The "KPa" in Table 7 should be "kPa"
Response 12
Thank you for your comment.
It has been corrected
Point 13
It is advisable to revise Figure 8 to make the graph easier to understand and the axis ranges need to be modified to fit the data.
Response 13
Thank you for your advise
Point 14
From Figure 10, it seems that the element erosion method was used, but the reviewer did not find a description in the manuscript, please explain.
Response 14
Thank you for your comment.
It has been added
Point 15
If the element erosion method was used, then the volume of the failed element can be calculated in LS-prepost, which would provide a quantitative analysis for this study.
Response 15
Thank you for your comment.
For this study, erode volume of elements after the simulation was not studied.
Point 16
The effective plastic strain in Figure 10 was essentially an index of damage, with strain hardening between 0 and 1 and strain softening between 1 and 2. Please explain this historical variable in the revision to avoid misinterpretation.
Response 16
Thank you for your comment.
It has been upadated
Point 17
Scaled distances for different load cases should be calculated and provided.
Response 17
Thank you for your comment.
It has been added in section 4.1.
Point 18
Is it possible to obtain advice on which structural forms and dimensions are safe at certain blast distances under the most adverse blast loading scenario?
Response 18
Thank you for your comment.
From this study, it is confirmed that when the thickness of the protection wall increased to 200 mm the structure gets low damage for both reinforced and composite structure at the distance of 3.6 m. Future study should be continue.
Point 19
The reinforcement arrangement in the RC wall needs to be supplemented.
Response 19
Thank you for your comment.
It has been corrected.
Author Response File: Author Response.pdf
Reviewer 3 Report
The manuscript investigates the behavior of barrier walls subject to explosions. The research is consistent with the aims and scopes of the journal.
The manuscript is very well written and I did not find any particular issue that should be addressed.
Possibly, authors could consider a better representation of the displacements in Figure 8 and reduce the font size of colorbars in Figure 10 (if possible). Nevertheless, such minor issues do not represent real drawbacks and could be implemented during the publication phase.
In conclusion, I recommend it to be accepted in the present form.
Author Response
Response to Reviewer 3 Comments
The manuscript investigates the behavior of barrier walls subject to explosions. The research is consistent with the aims and scopes of the journal.
The manuscript is very well written and I did not find any particular issue that should be addressed.
Possibly, authors could consider a better representation of the displacements in Figure 8 and reduce the font size of colorbars in Figure 10 (if possible). Nevertheless, such minor issues do not represent real drawbacks and could be implemented during the publication phase.
In conclusion, I recommend it to be accepted in the present form
Thank you for review
Point 1
Possibly, authors could consider a better representation of the displacements in Figure 8 and reduce the font size of colorbars in Figure 10 (if possible).
Response 1
Thank you for your comment.
Author Response File: Author Response.pdf
Reviewer 4 Report
1. There was no text describing Table 1. Please add a paragraph discussing the differences in the requirement of barrier wall in different countries. Also please add references for safety code requirement in Japan and in France.
2. The heading in Table 1 " Minimum requirement of barrier wall" would be better understood if it is changed to "Minimum requirement of barrier wall thickness."
3. Line 98 to 103: It is quite confusing the way it is now in the manuscript because most of the description is about the accident at hydrogen station in Gangneung, but in the middle of this, but another hydrogen at Samcheok is briefly described. It is suggested that the sentence is split into two sentences to separate the two hydrogen stations as follows: Figure 1 (a) shows the hydrogen tank explosion that occurred in Gangneung, South Korea, on May 2019. Figure 1 (b) shows the site visit at the hydrogen station at Samcheok, South Korea.
3.1 Then move the following sentences that describe Figure 1(b) to line 108 by inserting them in front of "it is necessary... " These sentences are"Figure 1 (b) shows the site visit at the hydrogen station at Samcheok, South Korea. The thickness of the reinforced concrete is 0.35 m in figure 1 (b), but the distance between the hydrogen gas trailer and the protection wall is 1.53 m. The distance between 102 the protection wall and hydrogen gas storage follows the International Fire Code (IFC) 2009 (States, 2007) [17]."
3.2 Then please describe the purpose of showing the picture of the Samcheok station next to the Gangneung station. Is it because they are similar in design, or something else?
4. In the text starting from Lines 207 to 212, Figure 3 was mistyped as Figure 2.
5. Line 190-191: What is X in Expression 1? How is its value determined?6. Line 195: Please provide a justification for choosing the value of The explosion efficiency of 0.01.
7.The sentence is broken from Line 209 to 210, "After the positive
phase the pressure becomes smaller than the ambient pressure and at
again returns to it called negative phase." Please rewrite with two or more sentences, it will be easier to understand.
8. Line 189: For better terminology, it is suggested that the word "ratio" is changed to " volumetric concentration" so that it would read " a 75 % volumetric concentration of hydrogen with air.
9. Line 295: Remove the word "of."
10. Line 255: What is Malvar? Is there a reference for this?
11. Table 6, third column: Change "form" to "from".
12. Line 315: Change "Table 6" to "Table 7."
13. Line 342-351: Figure 7 appears several times in the paragraph. Should the correct figure be "8" instead of "7."
14. Is it correct to say that Section 4.2 provides displacement results without considering failure from a detonation of 69.7 kg of TNT. And, the failures are discussed in section 4.3 based on the same simulation results as in Section 4.2.
If that is the case, please provide the failure criteria used to produce results in Section 4.3.
15. In figure 10, plots of plastic strain are shown according to color. Please add description on the meaning of cells that have no color.
16. Line 355-357, 376-377: The position of TNT was 0.55 m, but the detonation point was 0.75 m. Why the two position are different? Please describe the physical reason behind this.
Author Response
Response to Reviewer 4 Comments
Point 1
There was no text describing Table 1. Please add a paragraph discussing the differences in the requirement of barrier wall in different countries. Also please add references for safety code requirement in Japan and in France
Response 1
Thank you for your comment.
It has been corrected.
Point 2
The heading in Table 1 " Minimum requirement of barrier wall" would be better understood if it is changed to "Minimum requirement of barrier wall thickness."
Response 2
Thank you for your comment.
It has been corrected.
Point 3
Line 98 to 103: It is quite confusing the way it is now in the manuscript because most of the description is about the accident at hydrogen station in Gangneung, but in the middle of this, but another hydrogen at Samcheok is briefly described. It is suggested that the sentence is split into two sentences to separate the two hydrogen stations as follows: Figure 1 (a) shows the hydrogen tank explosion that occurred in Gangneung, South Korea, on May 2019. Figure 1 (b) shows the site visit at the hydrogen station at Samcheok, South Korea.
Response 3
Thank you for your comments.
It has been corrected.
Point 3.1
Then move the following sentences that describe Figure 1(b) to line 108 by inserting them in front of "it is necessary... " These sentences are"Figure 1 (b) shows the site visit at the hydrogen station at Samcheok, South Korea. The thickness of the reinforced concrete is 0.35 m in figure 1 (b), but the distance between the hydrogen gas trailer and the protection wall is 1.53 m. The distance between 102 the protection wall and hydrogen gas storage follows the International Fire Code (IFC) 2009 (States, 2007) [17]."
Response 3
Thank you for your comments.
It has been corrected.
Point 3.2
Then please describe the purpose of showing the picture of the Samcheok station next to the Gangneung station. Is it because they are similar in design, or something else?
Response 3
Thank you for your comments.
It has been added.
Point 4
In the text starting from Lines 207 to 212, Figure 3 was mistyped as Figure 2.
Response 4
Thank you for your comment.
It has been corrected.
Point 5
Line 190-191: What is X in Expression 1? How is its value determined?
Response 5
Thank you for your comment.
It has been corrected.
Point 6
Line 195: Please provide a justification for choosing the value of The explosion efficiency of 0.01.
Response 6
Thank you for your comment.
It has been corrected.
Point 7
The sentence is broken from Line 209 to 210, "After the positive
phase the pressure becomes smaller than the ambient pressure and at
again returns to it called negative phase." Please rewrite with two or more sentences, it will be easier to understand.
Response 7
Thank you for your comment.
It has been corrected
Point 8
Line 189: For better terminology, it is suggested that the word "ratio" is changed to " volumetric
Response 8
Thank you for your comment.
It has been corrected.
Point 9
Line 295: Remove the word "of."
Response 9
Thank you for your comment.
It has been corrected.
Point 10
Line 255: What is Malvar? Is there a reference for this?.
Response 10
Thank you for your comment.
Reference has been added.
Point 11
Table 6, third column: Change "form" to "from".
Response 11
Thank you for your comment.
It has been corrected.
Point 12
Line 315: Change "Table 6" to "Table 7."
Response 12
Thank you for your comment.
It has been corrected
Point 13
Line 342-351: Figure 7 appears several times in the paragraph. Should the correct figure be "8" instead of "7."
Response 13
Thank you for your comment.
It has been corrected.
Point 14
Is it correct to say that Section 4.2 provides displacement results without considering failure from a detonation of 69.7 kg of TNT. And, the failures are discussed in section 4.3 based on the same simulation results as in Section 4.2.
Response 14
Thank you for your comment.
Yes, section 4.2 and section 4.3 are based on the same simulation results.
Point 15
In figure 10, plots of plastic strain are shown according to color. Please add description on the meaning of cells that have no color.
Response 15
Thank you for your comment.
It has been corrected.
Point 16
Line 355-357, 376-377: The position of TNT was 0.55 m, but the detonation point was 0.75 m. Why the two position are different? Please describe the physical reason behind this.
Response 16
Thank you for your comment.
It has been corrected.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Thank you for the answers.
I still disagree with your answer to the point 1. I undrestand that you
would like to maximise the amount af TNT by increasing the hydrogen content to 75 %. The problem is that a significant amount of hydrogen gas will be left after explosion due to a very rich mixture.
If your domain is open, than you would need an additional time for the remaining hydrogen be mixed with air, following by secondary explosion.
Please clarify this point.
Author Response
Thank you for review
I still disagree with your answer to the point 1. I undrestand that you would like to maximise the amount af TNT by increasing the hydrogen content to 75 %. The problem is that a significant amount of hydrogen gas will be left after explosion due to a very rich mixture.
If your domain is open, than you would need an additional time for the remaining hydrogen be mixed with air, following by secondary explosion.
Please clarify this point.
Thank you for review
Point 1
Response 1
Thank you for your comment.
In present of 4 % to 75 % ratio of hydrogen with air, explode can occur (from references). In this study maximum percent was taken for all cases. TNT equivalent weight method was used for this study. TNT explosions are detonations, while vapor cloud explosions are deflagrations i.e., the shape and the velocity of their respective blast waves are different. TNT equivalency method is straightforward, it remains widely used. It has been added in manuscript with references.
Figure 8 represents the pressure vs time curve and for this study 10 milliseconds simulation is sufficient for the simulations. Increasing the simulation time can erode the few concrete elements after 10 milliseconds but the process is too slow and displacements on structure has not a large difference.
For secondary explosion current study did not consider it in this analysis. Future analysis and experiment study should be done.
Author Response File: Author Response.pdf
