Prediction of Pre- and Post-Breakage Behavior of Laminated Glass Using a Phase-Field Damage Model
Round 1
Reviewer 1 Report
The authors attempted a very challenging simulation task, predicting the fracture of laminated glazing under impact. The results show promise and the technique used is novel, though it seems to have some limit in its applicability which will need to be explored further. The paper would be interesting to other researchers, as it explores possible applications of the technique and could inform further research.
There are just a few points the authors should address before publication:
- The design of the specimen was very specific, several layers were used and the make-up was not symmetric. Some rationale behind this choice should be provided. It is understood that more detail would be provided in reference 9, however the readers of this paper would benefit from a brief summary, especially as the experimental paper has not been published yet.
- There is a typo in figure 3, it should be 7LG-4 at the bottom right.
- Ideally, all symbols should be defined within the paper at least quickly, rather than completely referring to other publications.
- There are also a few typos in the main text. For example, on line 107 I think the word should be weak, not week. Also, on lines 164 and 242 sentences seem to have been left which might not fit with the finished manuscript.
- Does the filter used for figure 12 also account for the longer decay of the experimental force?
- Has some form of hourglass control or fully integrated elements been tried out for the quadrilateral elements to avoid the zero energy oscillations discussed in line 218?
- When the simulations were run on the layered samples, for example 5LG-3 which is described in more detail in pages 15 and 16, was it attempted to simulate the repeated impacts, considering the existing damage on the layers for subsequent, higher energy simulations? This would ensure that the results are directly comparable to the experimental results.
- At the moment it seems that the initial capacity fin is a fitting parameter, which will need to be changed potentially for each layer. This is a serious drawback for the technique, as it would make its use difficult as a predictive tool for design or parametric studies. Do the authors have any suggestions about how the parameter could be determined in advance, if they think it is at all possible?
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
It is currently quite popular to simulate the crack propagation of materials and structures by using the phase field method. Here, the authors apply the phase-field method to the impact crack simulation of laminated glass plates. My comments for the manuscript are:
1. In addition to the phase field method, there are some characteristic methods that are effectively applied to crack propagation and impact simulation, such as the meshless method, the extended finite element method, the peridynamics, etc. It is recommended that the authors introduce the work related to these methods in the introduction and briefly analyze the advantages of using the phase field method in this paper.
2. How about the mesh-dependence of current phase field implementations on the modelling accuracy? And also the computational intensity.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Thank you for the work to improve the paper. I think my comments were dealt with adequately. I think only some minor revisions should be completed before publishing:
- I could not find the definition of the symbol beta in line 112. I might have missed it, but could it be added should it not be included?
- A quick proof read would be beneficial to smooth the English language. A further note to the writers was left in line 339
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
