Dear Dr. Okereke, Thank you again for the wonderful informative videos. I have a question, as we do tensile, compression, or shear tests on our RVE models we get some stress-strain values, however, these values can sometimes (or in most cases) not the same as experimental values. but the young modulus calculated from the slope shows a correct value. could you explain if this is normal? moreover, is there a way to calculate the stress-strain values that are same as the experimental ones? Thank you for your time and help.
Hello, this is a good comment. The numerical and experimental stress-strain plots (when the numerical ones are obtained by computational homogenization) should be the same. This is what I do here. However, you may not have them matching each other because of the underlying constitutive models of the fibre, matrix and interface in your model. You need all these properly represented for the experimental and numericla plots to match. If you want them to be correct, then get a correct matrix model. I usually use elastoplastic material model in my simulations, but real matrices do not behave this way. You might have to create a User-defined material model for your matrix which will capture the holistic behaviour of the matrix. With such, and adequate interface effects incorporated in the RVE model, then you have and comparable plot.
@@MichaelOkereke Thank you very much for your complete answer. Another question has arisen for me now, As I am a follower of all your videos I have realized that you mostly give mesh size of 10% of your RVE size, is there a reason behind it? Since I know we should do RVE size and Mesh dependent checking for our simulations before starting with the majority of them. Thank you again for your time and help.
Hi Sir I find your content very useful. I'm a final year engineering student. I need your help on how to simulate elastomers (made of Ecoflex) embedded with liquid metal (Gallium or its alloys) when I apply tensile deformation on it and change the phase of liquid metal by varying temperature.
Thanks for your commment and good luck with the research. I do not have any experience on the change of phase of liquid metal modelling. However, the framework I have published with my RVE modelling videos can work except you just have to specify the most appropriate material models for the constituents. I wish you luck with the research, but sorry I am not able to help right now.
Hello Sir, I have a question. I am working on a 3D unit cell model of a plate lattice structure. Because of the thin walls of the structure i modeled it with shell elements. Now i want to apply PBC to it but i am wondering how i have to constrain the rotational DOFs of the nodes on the oppsite faces.
Hello, thanks for your query. I am a bit confused that you are modelling a 3D unit cell and still somewhere combining it with a shell element. I hope this is right and you are correct in your description of the problem. I would have thought you either do a 3D (continuum) element problem or a 2D (shell or beam) element problem. Maybe, there is special case you are considering here. With regards to imposing PBC on the domain, this should be okay. Here is a video that I did a while back on PBC with 2D lattice structures: th-cam.com/video/4YDnKogZ_Ng/w-d-xo.htmlsi=2XMqpRV0CZmclIwJ This might help you. With respect to rotational degrees of freedom, I usually do not bother about the rotational degrees because the sort of mathematics of PBCs that I use is applied directly on nodes so a node-to-node coupling only allows for displacement degrees of freedom. Rotation DOF is not usually acceptable, as nodes do not show rotational degrees since they are points. If you are focussing on linking elements or segments of a beam, that rotational degrees of freedom will also be specified. I will adopt the same approach I use in defining the rotational DOFs as the directional displacement DOF. I hope I have not confused you but that is the thought I have about your problem.
I am modeling plate lattices similar to those shwon in the article: Additively-manufactured anisotropic and isotropic 3D plate-lattice materials for enhanced mechanical performance: Simulations & experiments My plates are very thin so i modeled them with S4R shell elements. The nodes of the shell elements do have in addition to the displacements the rotational DOFs UR1, UR2, UR3. I am wondering how to link these additional DOFs of two nodes on opposite faces to correctly apply the PBC. Thank you very mutch!
Hello sir, If I intend to model an explosion effect on the syntactic foam. How do you advice that I carry out the explosive modeling? Would it also be at a micro level?
Hello, I do not think a microscale modelling of the explosion is ideal. Normally, the sort of explosive modelling it at structural or macroscale format. This would usually, in ABAQUS, involves you using a CONWEP option under the interaction module to recreate the explosion. The length scales required is often very large and setting it up in the microscale is not feasible. There could be alternative ways, of directly applying a pressure wave on the microscale but I am not sure if there will be a sensible output from such a study. You may end up with localized stress rises and the deformation might not be representative of the macroscale explosive response. I am speculating here but my gut feeling is that I will not model this at microscale.
@@MichaelOkereke thank you so much for your prompt response. In line with your experience, how would you propose I model the foam at the macroscale level? Or perhaps if you have a video for it, I would love to know how to go about it. Thanks so much Dr. Michael.
Thank you sir for this work ! Please sir, i need your help and your advice i 'am student . Now, I am working on modeling fatigue behavior on my internship for parts from additive manufacturing, these parts contain lack of fusion defects (LoF) so I would like to model a 2D volume with abaqus then I will generate defects (with python) according to a distribution law (defect sizes, shape, etc.) then I apply a constraint on this volume in order to calculate the stress concentration and the fatigue life as well as the stress field distribution.
Hello, thanks for your query. I am not sure if I am able to help you with this query. I have no experience with Lack of Fusion defects specifically but could comment more broadly about the RVE modelling requirement. What the problem does sound to me is that you are considering isolating a 2D RVE of a additive manufactured material and want to incorporate some defects on the part and subsequently assess the effect of the defect on the macroscale (stress-strain) properties of the product/part. This makes sense as I can see the need for the research. It will be optimal to use numerical methods to assess the viability of 3D printed parts to help adjust the manufacturing process, if possible. If the above is correct, then I think there are a lot of videos on the CM Videos channel which I have about, and which can help you. I suggest you look at the videos about: 1. Creating virtual domains with microscale defects (look at videos on porous or void media) 2. Applying periodic boundary conditions on such heterogeneous domains (automatic and manually) 3. Extract homogenized effective properties (nearly every RVE modelling video, I always end with this) If you check these free resources and still unsure how to proceed and want to consult with me specially on this, then you can book a consultation with me via this link: calendly.com/michael-okereke/deep-dive-into-your-computational-modelling-challenge.
Thank you so much for your work! Thank you very much for your video!
You are welcome.
Dear Dr. Okereke,
Thank you again for the wonderful informative videos.
I have a question,
as we do tensile, compression, or shear tests on our RVE models we get some stress-strain values, however, these values can sometimes (or in most cases) not the same as experimental values. but the young modulus calculated from the slope shows a correct value. could you explain if this is normal?
moreover, is there a way to calculate the stress-strain values that are same as the experimental ones?
Thank you for your time and help.
Hello, this is a good comment.
The numerical and experimental stress-strain plots (when the numerical ones are obtained by computational homogenization) should be the same. This is what I do here. However, you may not have them matching each other because of the underlying constitutive models of the fibre, matrix and interface in your model.
You need all these properly represented for the experimental and numericla plots to match. If you want them to be correct, then get a correct matrix model. I usually use elastoplastic material model in my simulations, but real matrices do not behave this way.
You might have to create a User-defined material model for your matrix which will capture the holistic behaviour of the matrix. With such, and adequate interface effects incorporated in the RVE model, then you have and comparable plot.
@@MichaelOkereke Thank you very much for your complete answer.
Another question has arisen for me now,
As I am a follower of all your videos I have realized that you mostly give mesh size of 10% of your RVE size, is there a reason behind it?
Since I know we should do RVE size and Mesh dependent checking for our simulations before starting with the majority of them.
Thank you again for your time and help.
Hi Sir I find your content very useful. I'm a final year engineering student. I need your help on how to simulate elastomers (made of Ecoflex) embedded with liquid metal (Gallium or its alloys) when I apply tensile deformation on it and change the phase of liquid metal by varying temperature.
Thanks for your commment and good luck with the research. I do not have any experience on the change of phase of liquid metal modelling. However, the framework I have published with my RVE modelling videos can work except you just have to specify the most appropriate material models for the constituents. I wish you luck with the research, but sorry I am not able to help right now.
Hello Sir,
I have a question. I am working on a 3D unit cell model of a plate lattice structure. Because of the thin walls of the structure i modeled it with shell elements. Now i want to apply PBC to it but i am wondering how i have to constrain the rotational DOFs of the nodes on the oppsite faces.
Hello, thanks for your query.
I am a bit confused that you are modelling a 3D unit cell and still somewhere combining it with a shell element. I hope this is right and you are correct in your description of the problem. I would have thought you either do a 3D (continuum) element problem or a 2D (shell or beam) element problem. Maybe, there is special case you are considering here.
With regards to imposing PBC on the domain, this should be okay. Here is a video that I did a while back on PBC with 2D lattice structures: th-cam.com/video/4YDnKogZ_Ng/w-d-xo.htmlsi=2XMqpRV0CZmclIwJ
This might help you.
With respect to rotational degrees of freedom, I usually do not bother about the rotational degrees because the sort of mathematics of PBCs that I use is applied directly on nodes so a node-to-node coupling only allows for displacement degrees of freedom. Rotation DOF is not usually acceptable, as nodes do not show rotational degrees since they are points. If you are focussing on linking elements or segments of a beam, that rotational degrees of freedom will also be specified. I will adopt the same approach I use in defining the rotational DOFs as the directional displacement DOF.
I hope I have not confused you but that is the thought I have about your problem.
I am modeling plate lattices similar to those shwon in the article:
Additively-manufactured anisotropic and isotropic 3D plate-lattice materials for enhanced mechanical performance: Simulations & experiments
My plates are very thin so i modeled them with S4R shell elements. The nodes of the shell elements do have in addition to the displacements the rotational DOFs UR1, UR2, UR3. I am wondering how to link these additional DOFs of two nodes on opposite faces to correctly apply the PBC.
Thank you very mutch!
Hello sir,
If I intend to model an explosion effect on the syntactic foam.
How do you advice that I carry out the explosive modeling?
Would it also be at a micro level?
Hello, I do not think a microscale modelling of the explosion is ideal. Normally, the sort of explosive modelling it at structural or macroscale format. This would usually, in ABAQUS, involves you using a CONWEP option under the interaction module to recreate the explosion. The length scales required is often very large and setting it up in the microscale is not feasible.
There could be alternative ways, of directly applying a pressure wave on the microscale but I am not sure if there will be a sensible output from such a study. You may end up with localized stress rises and the deformation might not be representative of the macroscale explosive response. I am speculating here but my gut feeling is that I will not model this at microscale.
@@MichaelOkereke thank you so much for your prompt response.
In line with your experience, how would you propose I model the foam at the macroscale level? Or perhaps if you have a video for it, I would love to know how to go about it.
Thanks so much Dr. Michael.
Thank you sir for this work ! Please sir, i need your help and your advice i 'am student . Now, I am working on modeling fatigue behavior on my internship for parts from additive manufacturing, these parts contain lack of fusion defects (LoF) so I would like to model a 2D volume with abaqus then I will generate defects (with python) according to a distribution law (defect sizes, shape, etc.) then I apply a constraint on this volume in order to calculate the stress concentration and the fatigue life as well as the stress field distribution.
Hello, thanks for your query. I am not sure if I am able to help you with this query. I have no experience with Lack of Fusion defects specifically but could comment more broadly about the RVE modelling requirement.
What the problem does sound to me is that you are considering isolating a 2D RVE of a additive manufactured material and want to incorporate some defects on the part and subsequently assess the effect of the defect on the macroscale (stress-strain) properties of the product/part. This makes sense as I can see the need for the research. It will be optimal to use numerical methods to assess the viability of 3D printed parts to help adjust the manufacturing process, if possible.
If the above is correct, then I think there are a lot of videos on the CM Videos channel which I have about, and which can help you. I suggest you look at the videos about:
1. Creating virtual domains with microscale defects (look at videos on porous or void media)
2. Applying periodic boundary conditions on such heterogeneous domains (automatic and manually)
3. Extract homogenized effective properties (nearly every RVE modelling video, I always end with this)
If you check these free resources and still unsure how to proceed and want to consult with me specially on this, then you can book a consultation with me via this link: calendly.com/michael-okereke/deep-dive-into-your-computational-modelling-challenge.