Colored keys that represent breaks
Using the BPM package, I seem to have noticed that the package doesn’t handle the 2D mesh in a single-connected way, resulting in all six bonds of an atom breaking in the triangular lattice
Is it a statement or a question? I have no idea of which kind of advice you are expecting from posting a picture with no explanation.
I don’t think it’s right to interact with each other as long as the bonds of one atom are broken，So I wonder if it would be nice to add something like biconncomp in the graph in matlab
I’m confused, if your image is rendering broken bonds, what do you mean by “all six bonds break?” It looks like most atoms only have two broken bonds along a crack.
I don’t think it’s right to interact with each other as long as the bonds of one atom are broken，So I wonder if it would be nice to add something like biconncomp in the graph in matlab
How are atoms incorrectly interacting? With bond forces, pair forces? Is your concern that the implementation of the model you’re using (bpm/spring, bpm/rotational?) is incorrect, or are you requesting a new model be added? I’m not familiar with Matlab, so the analogy is unfortunately lost on me.
To get more helpful responses, I recommend reading the guidelines for posting. I’d like to help, but it’s hard to give good advice without a clear question and system details.
Hello Author,
I am using a triangular lattice where each atom has six bonds. When rendering the fractured bonds, I noticed that some atoms have all six bonds broken. For instance, when I highlight these special points after rendering and statistically analyze around 100 samples, it frequently happens that all six bonds of an atom are broken. My point is, mechanically, I believe that an atom should be able to break at most five bonds. When only one bond remains, as shown by the red bond number one below, I hypothesize that this bond should not break. I suspect that there may be numerical errors leading to the breakage of the sixth bond.
Yesterday, I discovered a method in C language using graph theory. For example, when structures 6, 7, and 8 are connected to the main structure by only one bond, this method can be used to identify them to prevent all bonds of an atom from breaking.
7.212.data (456.6 KB)
Thank you for the extra detail, I think I understand your question now.
I’m not convinced this implies there is a bug in the code. If one particle suddenly had a large impulse (due to, say, being located at a crack front), why shouldn’t it be possible that it breaks all bonds? At some temperature the system will melt and all bonds will break.
If you think otherwise, I recommend finding an example when one bond loses all 6 bonds (say outputting when bonds break using store/local) then rerunning and regularly outputting dump files near that instance in time. If you find a timestep where a bond incorrectly breaks without reaching the threshold strain, then that would be a good indication of numerical errors and would be a good example for debugging.
In my case of quasi-static loading, I have set energy minimization at each step to achieve the optimal energy state. During this process, the effect of temperature is not considered.I read in one of your articles that when the tensile rate is very low, it can achieve static loading.If static tensile conditions can be achieved, does that mean the effect of temperature does not need to be considered?
My recollection is that the BPM package doesn’t calculate bond energies since the damping is dissipative, so I wouldn’t expect energy minimization to work.
However, I still think observing atom trajectories near one of these events you think is faulty might be the most informative.