Ahmed Elkady really looks like you are an engineer with solid knowledge in both the mechanics of materials and the Abaqus software. I really appreciate the knowledge you share. I am convinced that in some way: life, karma (destiny), God, etc; It will return in blessings the good will you have to teach and share what you know. I send you a big hug from Peru (this is a way to say an affectionate farewell when you say goodbye to someone). Sorry if the translation is not good, I used a translator.
Please note that the proper way to compute the plastic strain is to subtract sigma@current strain/E from the total strain. This is actually the method used in the EXCEL file provided in the link in the video description. However, sometimes for simplicity, you can just subtract sigma_yeild/E; the difference is generally minor for steel materials with relatively large E.
Good explanation. I've an observation about what you said when you defined the engineering strain: L0 is the length of the specimen's uniform section named "gauge length". If you consider this uniform section, between the fillets, you have an uniaxial stress state in each point. So the engineering-stress strain curve is correctly defined if and only if we consider these hypotesis: - uniaxial stress state; - small displacement; - volume conservation; Other curves: true stress-strain from elongation and true stress-strain from section, are defined by the violation of the cited hypotesis.
Thank's, Dr. Ahmed please can you give me answer for my question ? When analyzing metal fatigue, should we prioritize the first plastic strain or the equivalent plastic strain? How might this choice impact the accuracy and efficiency of our results?
I've got a question. It seems in the Abaqus documentation that the plastic strain must be evaluated not only on the basis of the initial yield stress. You always remove SigmaYield/E whereas this should be updated as SigmaTrue/E at each SigmaTrue level according to the doc. Do you have any point of view on that? This is a difference of understanding of the method between several of my colleagues.
Correct, the proper way to compute the plastic strain is to subtract sigma@current strain/E from the total strain. This is actually the method used in the EXCEL file provided in the link in the video description. However, sometimes for simplicity, you can just subtract sigma_yeild/E; the difference is generally minor for steel materails that have relatively large E.
In case, if we use a strain gauge attached to a coupon sample for measuring the strains directly ,,does this strain is considered as Engineering or true strain ???
Sir I'm doing CFST column steel beam joint under cyclic loading. My model is running with displacement at the beam end But it is showing convergence error when it is subjected to cyclic loading at beam end Will it be because of any error in material property given? Can u please share ur view on the topic Please help
Dear Mr. Elkady, I was wondering how to get the data of plasticity if we use the temperature-dependent data? I am confused on how to pick the data. Because the strain curve for each temperature differs. Thank you in advance and looking forward for the answer
Thermally-dependant plasticity can be defined through the Johnson-cook mode. There are many publications online describing the model parameters calibration and definition.
Thank you sir. But I have promblem. I am working on V bending 2D plain strain. When I define true stress- true strain graphs all points for plasticty in the material definition, the results of the springback are less than the experiment. However, if I define only the line where the plastic strain is 0 (the first line (yield point)) as you describe in the video 19:46 , the spring-back estimation gets much closer to the experimental results. Is it a correct method to define only the line where the plastic strain is 0? Do you have any other suggestions to increase springback in terms of material? I should point out that; the material is copper and I am using stress strain curves from other articles. Because the experiment I'm analyzing is not mine.
You should model the material's elasto-plastic behaviour to match the material uniaxial tensile test (true stress-strain curve). Perhaps a bilinear elastic-perfectly plastic model matches your material the best.
Thank you sir, great video and explanation. Just one question: I think the plastic deformation is not difference between total deformation and elastic deformation for yield stress but it is difference between total deformation and elastic component of deformation for the current stress. In video time 16:33 equation "Epsylon_pl = Epsylon_true - Epsylon_y" should be "Epsylon_pl = Epsylon_true - Epsylon_el" => "Epsylon_pl = Epsylon_true - Sigma_true/E". Am I right? If not please explain.
Yes, the proper way to compute the plastic strain is to subtract sigma@current strain/E from the total strain. This is actually the method used in the EXCEL file provided in the link in the video description. However, sometimes for simplicity, you can just subtract sigma_yeild/E; the difference is generally minor for steel materials with relatively large E.
What if we only have engineering stress and not true stress at fracture? In such case we can not calculate the elastic component at the fracture as we lack true stress and ultimately can not calculate plastic strain at material failure. What do we do then? Any suggestions will be appreciated. Thank you.
Hello, I don't quite understand why your plastic strain is minus a constant? Shouldn't the elastic strain be changing all the time?Different stresses correspond to different elastic deformations
Correct, the elastic strain value will change with stress. This is how the plastic strain is dedcued in the referenced Excel sheet on GitHub. Sometimes however, it can be acceptable to just subtract the yield strain for simplicity.
Oh no. I just realized that I've made simulations with engineering stress-strain graphs so far. Oops... Thanks for enlightening me on this mistake of mine.
Hello sir I want to ask you simple question.. I finished my simulation in abaqus of tensile test now i put a script in python to allow me to change value of young's modulus 10 times si i've now 10 inp and odb files so all what I want is how can I call abaqus to read this odb files automatically to exctract the graph of stress strain or if there is a simple way to do this
Hi, your video was awesome. If i plot in "plastic table" only 1 data like your example of MIN 19:40 I will design in elastic domain right? Please can you help me with this doubt.
No, in that case, you have defined an elastic perfectly-plastic material. No need to define a plastic behaviour if you are conducting an elastic analysis.
Thanks for the great explanation. I have a question if you don't mind. If I'm validating an experiment using ABAQUS, the author didn't provide a stress-strain curve he just mentioned the yield and ultimate stresses and their corresponding strain as a result a tensile coupon test. Should I assume that the mentioned points are the true stress-strain points ? And put them directly to ABAQUS?
Commonly, researchers report the engineering values not the true ones; so I would assume its the former. One more way to check is by comparing the reported ultimate stress with the expected value that is reported in the design standards for the same material grade (which is also an engineering one).
Yes, it should be true stress-strain. But note that it probably does not have significant impact for rebar; actually, many just model rebar using elastic-perfectly-plastic material.
Great video thanks. I have a question if you do not mind. I see that true fracture stress is above the ultimate stress. However, in abaqus documentation, the stress is lower than the ultimate value due to damage accumulation. 130.149.89.49:2080/v6.14/books/usb/default.htm?startat=pt05ch23s02abm18.html (Figure 24.2.3-1) Assuming that I am comparing with physical test, which one is correct? I am bit confused here.
The true stress of metalic materials such as steel increases with increasing strain. The figure you are refering to is concerened with the phenomenological ductile damage model which is a material-property-add-on that degrades the material's undamaged stress (discussed in this tutorial) to simulate strain softening behaviour leading to fracture.