Clarification Request: Interblock Critical Values in Damage Models

[franciscoyapor]

Hello PeriLab Developers,

I’ve been exploring the PeriLab.jl framework and came across the configuration for interblock critical values in the damage models. Here’s an example from one of the test cases: ./test/fullscale_tests/test_critical_stretch/critical_stretch_interface.yaml

Damage Models:
  Dam_1:
    Damage Model: "Critical Stretch"
    Critical Value: 0.1
    Interblock Damage:
      Interblock Critical Value 1_2: 0.2
      Interblock Critical Value 2_1: 0.1

I would appreciate clarification on the following:

1. Physical Meaning: What is the physical significance of the "Interblock Critical Values"? How do they relate to the failure mechanics between different material blocks?

2. Direction-Specific Values: In the example above, Interblock Critical Value 1_2 and Interblock Critical Value 2_1 are different. Does this mean a bond originating in block 2 fails earlier than one originating in block 1? Could you elaborate on how this asymmetry is interpreted and in what scenarios such values would be relevant?

3. Critical Energy Release Rate Damage Model: How do the interblock critical values work for the critical energy release rate damage model? I imagine it would differ from the critical stretch model since the physics behind the failure mechanism is different. Could you explain how interblock values relate in this context?

4. Application to Multi-Material Models: How would one correctly apply interblock critical values when modeling a system with multiple materials? For example, if two blocks represent materials with significantly different properties, how should the interblock critical values be set for either the critical stretch or critical energy release rate models? Any practical guidance would be greatly appreciated.

Thank you for your help and for providing this robust tool for peridynamic modeling!

Best regards,
Francisco

[CWillberg]

Hello @franciscoyapor,

I try to answer your questions.

1. We gave this option to allow some form of interface behavior. Only bonds which runs over block boundaries are affected. If you apply the same damage model as for the bonds inside the block, nothing happens.
2. As everything in PD it is directions depended. Assume block_2 has no damage criterion, the interblock won't be analysed. The same is true for other damage models as well. Also if you do not use geometry based failure models as critical stretch, you get a direction dependency as well.
3. Yes both models are different. For the interblock damage, basically nothing changes. Assume all critical values inside a block are value A and for the interblock its value B. The alternative form my point of view is to increase the resolution and apply a new block. Its not exactly the same, but hopefully you get the idea. Also you can apply transformations there, to approximate some kind of joint with anisotropic behavior.
4. Thats a good example of the idea. If you do not have the interblock option all interblock bonds have the paramter from material A in one direction and material B in the other direction. We did this and got interface damages. However, the interblock damage allows it to define for both bond (12 and 21) the same critical value, if you like. I do not know the correct value. It might be something to study. If you are interessted, we can discuss a numerical study about it and write a paper.

The point is that most of the papers do not deal with this problem, because homogeneous material is used. Therefore, it is open for research and discussion.

Hopefully this answer helps you.
Christian

[franciscoyapor]

Hello @CWillberg ,

Thank you for your detailed response and for clarifying the behavior of interblock critical values. Your explanation really helped me understand how interblock bonds work and how they enable interface behavior modeling. I also appreciate your insights into the differences between the critical stretch and critical energy release rate models and how these concepts apply to multi-material systems.

My research focuses on damage propagation in functionally-graded materials, where interface behavior is a key area of interest because of the material property gradients. The ability to model anisotropic or direction-dependent interface damage is directly related to the challenges I’m working on. Your suggestion to conduct a numerical study and explore appropriate interblock critical values sounds very exciting. I agree that this could lead to valuable contributions, especially since, as you mentioned, most studies tend to focus on homogeneous materials.

If you and @JTHesse are interested, I’d love to contribute and collaborate on this topic using PeriLab.jl. Maybe we could start by discussing ideas for a numerical study, including setups and potential material systems to focus on? I’m happy to exchange ideas over email if that’s more convenient for a deeper discussion.

Let me know how you’d like to move forward. Thanks again for your insights and for considering this as a potential research project!

Best regards,
Francisco

[CWillberg]

Hello Francisco, we are interested in collaborations. If you like we can organize a zoom meeting and discuss some ideas. Additions to the code or documentations are always welcome. Christian Von: Francisco A. Yapor Genao ***@***.***> Gesendet: Montag, 27. Januar 2025 06:50 An: PeriHub/PeriLab.jl ***@***.***> Cc: Christian Willberg ***@***.***>; Mention ***@***.***> Betreff: Re: [PeriHub/PeriLab.jl] Clarification Request: Interblock Critical Values in Damage Models (Discussion #221) Hello @CWillberg <https://github.com/CWillberg> , Thank you for your detailed response and for clarifying the behavior of interblock critical values. Your explanation really helped me understand how interblock bonds work and how they enable interface behavior modeling. I also appreciate your insights into the differences between the critical stretch and critical energy release rate models and how these concepts apply to multi-material systems. My research focuses on damage propagation in functionally-graded materials, where interface behavior is a key area of interest because of the material property gradients. The ability to model anisotropic or direction-dependent interface damage is directly related to the challenges I’m working on. Your suggestion to conduct a numerical study and explore appropriate interblock critical values sounds very exciting. I agree that this could lead to valuable contributions, especially since, as you mentioned, most studies tend to focus on homogeneous materials. If you and @JTHesse <https://github.com/JTHesse> are interested, I’d love to contribute and collaborate on this topic using PeriLab.jl. Maybe we could start by discussing ideas for a numerical study, including setups and potential material systems to focus on? I’m happy to exchange ideas over email if that’s more convenient for a deeper discussion. Let me know how you’d like to move forward. Thanks again for your insights and for considering this as a potential research project! Best regards, Francisco — Reply to this email directly, view it on GitHub <#221 (reply in thread)> , or unsubscribe <https://github.com/notifications/unsubscribe-auth/AIHBTC7PW2TVFLDSO33OZED2MXCJHAVCNFSM6AAAAABVRXNNSKVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTCOJWGQ4DOOI> . You are receiving this because you were mentioned.Message ID: ***@***.***>