It is not available yet. Watch this space!

Was going to ask this as well. Because I believe equivalent plastic displacement is then used to get from this value to the point of failure in the damage evolution module.

Hi, good question. If you look at the graph, the initiation aof damage is at the UTS and there D = 0, which suggests that damage has started and this will evolve with changing D until D = 1 (complete damage). Complete damage occurs at the fracture/separation of the specimen. The strain at which this happens is the fracture strain. It could not be at the epsilon_0 (which is equivalent to start of the plastic strain), as this corresponds to the UTS. The specimen cannot fracture at this point rather it will be the final point of yielding. Eventual fracture/separation of the specimen happens at what I have called fracture strain. Within a single element, when damage initiates, we need to know when a failed element will be deleted from the model. This corresponds to the fracture strain through the displacement at failure. Once this is obeyed, the element is then deleted from the model since element deletion is swithced on. Other elements around the failed element might have D values not up to 1 and so not approached the fracture strain. Once they meet this condition, the elements would then fracture and be deleted from the the model. If you do not believe me, I asked ChatGPT to define fracture strain and here is the answer: "In the context of continuum damage mechanics, fracture strain is defined as the strain at which a material undergoes fracture, leading to the complete loss of load-carrying capacity in the material. It is The strain value at which a material experiences fracture due to the accumulation of damage, resulting in a complete loss of its structural integrity." I believe this is consistent with what I have tried to explain above.

@@MichaelOkereke Thank you sir, that's very clear. But in Abaqus manual Ductile Damage Section, the "Fracture Strain" is defined as Equivalent fracture strain at damage initiation, not the end. I tested it in abaqus and plotted the stress-strain curve, and the "Fracture Strain" input value is really just the initial failure strain.

Yes you can! I will think about the thermal analysis aspect.

Yes you can! I will think about the thermal analysis aspect.

Hello, please send me an email on the email address in the About section of this channel or on [email protected]

You're welcome!

If it is not clear from my video, I suggest you google how to do it and there may be more guide on the ABAQUS documentation. It is difficult to offer support on this comments section. Good luck with the work.

@@MichaelOkereke thanks

Good question! Lode parameter is definitely essential for describing plasticity of materials. I have not specified it here as the ductile damage model that comes with ABAQUS, and which was reviewed in this video does not have need of this lode parameter. Just like the stress triaxiality, the lode parameter gives and indication of the dominant stress state in the model. Lode parameter, typically represented by mu, can be: mu = 1 (pure shear stress), mu = 0 (uniaxial stress) and mu = -1 (corresponds to a stress state where principal stress, σ2 =σ3 (triaxial compression). It is similar to the stress triaxiality term which gives us an indication of the stress state (dominating) the simulation. If you want to learn a bit more about the lode parameter and the like for a triaxially loaded system, then watch this video of mine: kzbin.info/www/bejne/fnuWe62Gis2tZ7c).

Yes you can! That will be a truss-based 1D-analysis but yes you can.

I do not have such a video but could explore possibility of making it. I have not done such before but this should be an interesting one to work on.

@@MichaelOkereke Thank you for the kind reply Prof. Because hardness data is the basic understanding of the material's strength. if we relate this data and Comsol simulated data then it will be good findings before going to tensile or any other bulk characterization.

Good question. The reason is that in extracting poisson ratio, you need two displacement (or strain measurements) i.e. u1 (along direction of loading) and u2 (transverse to direction of loading). These need to be extracted at different points in the model i.e. u1 from the node in direction of loading and u2 (on a node at the y-axis, transverse direction of loading). As I have only one RP i.e. at U1, then that gives me the displacement there. To get that of U2, I can do it two ways: 1. Use field output to track that (as displacement is always available for every note in the model) OR 2. Introduce another RP at that node and track it similarly. I decided on the first one, as I wanted students to be able to extract poisson ratio easily with their normal tensile simulations as u1 will always be there. You can then go to an applicable node to extract u2 from a field output. This way, you do not have to specially setup the model for poisson ratio by introducing an RP node. I hope the explanation makes sense.

Glad it was helpful! Well done Sherry.

Hello, This is certainly possible. In fact ABAQUS documentation has an example which you can adapt for your case. Here is the link to the example: classes.engineering.wustl.edu/2009/spring/mase5513/abaqus/docs/v6.5/books/cmd/default.htm?startat=pt01.html You may have to consider this: from odbAccess import openOdb odb = openOdb('job_name.odb') I think it is possible to analyze .odb file directly from the ABAQUS command window. I am not sure about running it directly inside Python - but if so, then you do need to import the right object files, which it seems to me this is what is happening with your error. I suggest you run it inside the ABAQUS command window or the ABAQUS kernal window (located foot of ABAQUS CAE page). This way, ABAQUS already has the object files like the odbAccess, so you do not have to worry about them. If you run inside Python, you will have to worry about where they are and bring them in. I hope you are able to do it.

@@MichaelOkereke Thank you for your response Sir. Actually, I was able to extract the results and complete the entire process outside of Abaqus using just the external Python interpreter. I used the following command: command = sprintf('abaqus python extract_results.py %s', odbFile); [status, cmdout] = system(command); Instead of just "python extract_results.py", and everything worked perfectly

I am happy to see that it all worked out well. Well done and keep it up.

Hello Manish, You are right and here are my thoughts: 1. For systems that are not isotropic, then you would have to find an applicable equation for determining bulk modulus for such structures. If it is a composite, then the classical laminate theory comes to mind. For particulate systems, there are also equations for extracting effective properties for heterogeneous media. I will leave you to find this out. I simply showed how to think logically about this, and verify your numerical model using a simple isotropic model. For anything more complex, the student will have to solve that for themselves. 2. The simulation I have shown for BCC system is a unit cell simulation but not based on a periodic boundary condition. You will have to look at my videos on dealing with PBCs on lattice structures (albeit 2D) to understand. For 3D systems, you need a different ABAQUS plugin to implement true 3D PBC on your lattice structure. I have not done so here. Good luck with the research.

Hello, This is a good point. The .stt file is a state file used in ABAQUS to store information about the simulation at every timestep during the simulation. It is basically the workhorse of ABAQUS and captures not only the time-step simulation information associated with the simulation. For example, whilst the ODB file may store the results after a successful convergence within a time step, the .stt file will store every the iterations that led to a successful result. This is why, often for highly nonlinear simulations, the .stt file will be bigger than the ODB file. What you can do to reduce the memory load is to ask ABAQUS to only write to the .stt file at certain frequencies (where you demand for that output of interest). I did not include it in my list but certainly it can be. Infact, here is an extensive list of all possible file extension types in ABAQUS: classes.engineering.wustl.edu/2009/spring/mase5513/abaqus/docs/v6.6/books/usb/default.htm?startat=pt01ch03s05aus29.html

@@MichaelOkereke Dear Professor, then can I delete those stt files once I finish with my simulation, provided I don't need any restart analysis further?

Yes in deed

The simplest way is similar to what I did here but replace the volume of the cubic element by the volume of a tetragonal element, but still take the cube root of that volume. Take volume of a tetragon = a^2c where a = edge length of a regular square base of the tetragon and c is height. I think you could assume that the elements you are dealing with are regular tetragons even though in more cases it would be irregular but deviations from edge length, 'a' and 'c' can become an error measure which you can factor too in determining the characteristic length.

Hello, thanks for the query. This is possible. You simply need to identify a set of elements in the surface you want to. Associate them to a set and then use history variables to track them later.

@@MichaelOkereke for making the set i need to make a section in middle ? Or there is other ways.

Glad it was helpful! That is what my intention was and I am glad that vision is achieved. Cheers!

Not really...I was working in an SI unit with mass in a unit of kg. Also, the Pa is N/mm^2 and N = kg.mms^(-2). Therefore, it's better specified in kg. However, if your masses are in tonnes, then indeed you are right.

Thanks very much. I will keep going. Any suggestions for future videos?

I have a particular interest in modeling superplasticity of metals, maybe use some of your syntactic RVEs to investigate the effect of the voids on the overall behavior of the material!!

Yes, that is fine. I think you can also look at the video about Triaxiality - some of the ideas in that video might help you with the superplasticity argument. Well done!

You are right... it could be easily done with ABAQUS Implicit and it would work. However whenever there is significant possibility of nonlinearity, damage, impact or the examples you cited, it is usually recommended to include an ABAQUS Explicit step as its better suited to deal with the convergence of such unstable simulations. I hope it makes sense.

Good question. I keep planning to make such video. Please watch the space.

I had a thought about it and the simplest way is similar to what I did here but replace the volume of the cubic element by the volume of a tetrahedron shape, but still take the cube root of the volume of the tetrahedron. Take volume of a tetrahedron = a^3/(6*square-root(2)) where a = edge length of a regular tetrahedron. I think you could assume that the elements you are dealing with are regular tetrahedrons where in more cases it would be irregular but deviations from edge length, 'a' can become an error measure which you can factor in in determining the characteristic length.

Glad you enjoyed it and felt that way about it. I wanted to make a comprehensive video about ductile damage that tries to show the whole theory and simulation setup for viewers. Glad it struck a cord with you!

​@@MichaelOkereke Thank you for your great effort. I was wondering if the method you described can be applied to truss elements like T3D2 or beam elements like B31.

I will not think so, as the element type used here are 3d continuum elements so am not sure you can use it for the beam or truss elements.

@@MichaelOkereke I am wondering if there is a different approach. This would be significantly important when modelling actual-size reinforced concrete subjected to severe loading such as cyclic load. In this case, the rupture of the reiforcing bar could be identified at critical locations (plastic hinge locations). Anyway, thank you, Dr. Michael Okereke, for your reply

This is a good question. I have not done this before but should be interesting. What is the matrix and metallic reinforcement you are taling about. What is the hybrid material?

@@MichaelOkereke it may be aluminum matrix and TIB2 + SiC particle reinforcement. Analysis should be on damage initiation, crack propagation and interfacial debonding.

Most welcome

Okay, noted.

Thanks for the feedback!

I really do not know, sorry!

Meaning?

Appreciate it!

Please check my other videos on unit cell and see if you can find resources to help you.

Hello, Thanks for your query. Here are my answers: 1. I prefer to associate the constraint to geometries instead of nodal sets as the later is subject to meshing of the model. If you change the mesh, you will have to re-attach/re-impose the constraints on the new nodes. But when you associate the constraint to a geometry, since the geometry does not change with the simulation, when you re-mesh, it will still inherit the constraint from the previous mesh, since those constraints are attached the the geometry. This is why I use geometries instead of mesh nodal sets. 2. The reason for using the negative shear value is simply aesthetic - nothing technical. The idea is that with the shear pushing away from the viewer, you would see better the shear deformation rather than when it pushes towards the viewer. Hope this makes sense? Thanks.

I used linear patterning both vertically and horizontally.

Thanks.

Glad it was helpful!

I did already in this video, not sure if you saw it: kzbin.info/www/bejne/n3_Hd6JrrdyGqqs

@@MichaelOkereke Thank you for your reply. I have already seen that video, and in that video, you have only discussed elastic properties. Actually, I want to know about the ductile damage, fracture behavior, and hardening properties of metal matrix composite.

Not sure about this... sorry!

Here is the video you need at the 4:59 mark kzbin.info/www/bejne/n3_Hd6JrrdyGqqssi=Qv9J78UfrYSR7EeI

Thank you ​@@MichaelOkereke

You have to use interphase model probably with a cohesive zone element.

Hello, I do not know much about the embedded constraint option but I have seen it being used. I reckon if you use it, you would not have to use the merge function but do not quote me in this. It is something I need to research further myself to be able to tell you definitively.

Hello, I do not really understand the question. I suppose you want to learn how to write a Python code to create RVEs. If that is the case, then I am unable to help with this. I have a code that can create particulate systems, just look in the description section of my videos.

Good question Johnson, it depends on how you want to setup the model. The right approach is to just represent the material as robustly as posible using say the tensile properties. Then impose a bending load on the model. The combination of bending loads with tensile material properties will combine to give you the right kind of response. So, the simple answer is use the tensile properties, then impose bending load, then extract force displacement plot which then gives you the bending behaviour. Then, use the force-displacement data to be compared against the experimental bending data set (what we call validation). Hope this makes sense to you?

I have always had this idea about creating plate-like particles. I have thoughts of how it can be done very manually but I think the best approach will be to create it automatically using some kind of Monte Carlo method. This later is definitely hard and creating a python script to do it is not trivial. I will keep thinking about it and see if I can make such a video but that is a good question.