General Mechanical

General Mechanical

When should I use material Non-Linearity?

    • Rameez_ul_Haq
      Subscriber

      Most of the times, for an isotropic material, we are interested in knowing if a ductile material will plastically deform or not. If it does, then we say it has actually failed. So I just have a basic question like when should I be making use of the material's non-linear properties? Where can this help to be useful?

    • Erik Kostson
      Ansys Employee
      An example would be predicting the ultimate load a structure can take before it fails (say concrete slab). If we went by the load it takes to first crack (or first plastify). When designing structure, one might also want to find out what the ultimate load is before it fails, so we want to push beyond first cracking in concrete (or initial plasticity). ALso for buckling of structure the linear buckling might give very conservative results, while in reality if we include all of the nonlinear (nl) effects (including material nl) we might find the buckling load is different than the linear buckling prediction. Another example is the designing of connections in buildings where we need to use nl material, e.g., see this example for details :
      https://www.ideastatica.com/support-center/capacity-design-en

      So these are just a few examples, and sure there are many more.
      For the first mentioned above See an example of a nonlinear material analysis (concrete nonlinear material), where we predict the failure load of a slab (this analysis requires nonlinear material and large deflections).
      Erik

    • Rameez_ul_Haq
      Subscriber
      ,I am not very familiar with concrete materials since I haven't worked on it, but as far as my knowledge goes, concrete is strong in compression and not in tension. And the crack thing is most oftenly related to the materials in tension. I don't think the mechanism of crack propogation beginning at yield strength and causing a fracture past it, is eligible for the materials in compression (including concrete). I think a different mechanism causes a yielding and fracture for materials in compression, not the crack propogation and enlargening of the cracks. So I don't know why you related the concrete example (under compression) to the crack propogation and ultimate fracture.
    • Erik Kostson
      Ansys Employee
      HI ´╗┐@Rameez_ul_Haq´╗┐
      First all a "thank you" for people helping you out would
      be good a start for your posts, and much appreciated. Thank you. So kindly
      appreciate the support and the time I and my Ansys colleagues take to help you
      and other students in this forum where you do not pay for support.

      So please always have that in your mind , that we do this because
      of our good will to help you and anyone else in the forum. I am not obliged to
      help you and you are not entitled to support or help (that would only be the
      case if someone is paying for Ansys support, like some Ansys users do). So be
      thankful first and have this in your mind. Also formulate your questions
      better, as it is extremely general, like you said "when should I be making
      use of the material's non-linear properties". There could be 1000 of
      answers to this and the one I send you is one or two. So one should use nl material when predicting the failure and ultimate load capacity for a reinforced concrete slab answers that question.

      AS I said kindly also look at the other examples where one could
      calculate the ultimate load capacity for a connection - that requires a nl
      material analysis (I mentioned 3 not only concrete), and there are many more
      examples where nl material can be used to calculate ultimate load capacities.
      So the concrete is another example where nl material is used to calculate the
      load capacity of a structure, we do not to think what and how it behaves, it is
      just another form of material nonlinear behaviour (the concrete material we
      have do not really look at cracks they (material models) are based on
      plasticity, but that is not the point here). Other examples are metal forming where we form material which plasticizes of course, and many more manufacturing
      type of analysis need nl material. So just search there are many examples of
      where we need this.

      All the best, and good luck.

      Erik
    • Rameez_ul_Haq
      Subscriber
      thank you sir. I have always shown my gratitude to you and other ANSYS employees on any post or comment I make, for helping the students and engineers around the world to become better versions of themselves. This platform is a knowledge station, and It is your support and help that keeps this community thriving and getting their confusions and questions sorted out which aids them throughout their lives. Pardon me if you felt the reply was rude, it didn't mean to be :) I just had a click in my mind after reading your answer (based upon the knowledge that I already have), thats why I replied instantly.

    • Erik Kostson
      Ansys Employee
      Thank you for your reply. No worries - all the best for now.

      Erik
    • Rameez_ul_Haq
      Subscriber
      ,I would like to ask you one more thing, if you may :) I saw the link that you attached for determining the maximum load bearing capacity of a concrete slab. My question would be that the graph that you attached there, i.e. the graph of reaction force against time, so was the final value of the reaction force at 1 sec equal to the applied load? I mean I couldn't understand actually that why did the load suddenly decrease? If I am conducting a non-linear analysis, then I input a load and it is applied in steps until it becomes equal to the value that I have inputted at t = 1 sec. At each sub-step convergence, the reaction force is always somewhat equal to the value of external force at that sub-step. So how is it possible that the reaction force decreases suddenly. Does that mean that the applied load has decreased suddenly, or something else?
      I hope you have understood what I am trying to convey :)
    • Rameez_ul_Haq
      Subscriber
      ,if I can get an answer from you on this, that would be extremely helpful :) Thank you.
    • Erik Kostson
      Ansys Employee
      Good Morning

      Of course.

      In that analysis in order to predict failure in concrete material models (Menetry Willam material model), where it becomes very difficult to get convergence once the structure fails, we use an enforced displacement to push the slab. So what we plot there is the total reaction which is the total force applied on the slab.

      Now what happens when the force drops means that it can not take anymore load (so it drops) since the response/graph is going down (what we call a softening response of a structure - this is due to the strain softening in concrete), and this is due to the fact that the slab has failed in punching shear (large plastic zone marked in red second image below, shows that a plastic cone has formed similar to what you see in punching shear failure like in the first image below)
      ).

      So the largest force on that graph is the total failure load (so the largest reaction observed in the graph not its value at 1 s which is not relevant - only the largest reaction we need here) of the slab which in this case corresponded well with experimental tests. AFter that the structure softens due to the large plastic strain region showing a cone like shape just as in real experiments.
    • Rameez_ul_Haq
      Subscriber
      thank you for your kind reply. I think that the enforced displacement that you used comes into play as soon as the structure fails. For isotropic ductile materials, I think this should be done as soon as the necking of the material starts. Please tell me if I am right or wrong on this?
      Can you please also enlighten me that how would I know that now I should use something extra as an input (like you used an enforced displacement) so that the solver is able to converge after the failure has occurred? Or should we first conduct the analysis without that something extra, and check if the analysis is converging or not, and if it is not, then we need to put that something extra to make it converge?
      And one more thing, at t = 0.9 secs when the reaction force is the highest, does this mean that all the applied load is the reaction force there? Or still that is a % of the total applied load? (talking about the situation before the enforced displacement comes into play).
    • Rameez_ul_Haq
      Subscriber
      ,and also, if you may, let me share a video uploaded by ANSYS YouTube channel which properly illustrate how to model a multilinear plastic material properties.
      In this video, the data for the tensile strength test is inputted. But what if I have a structure which is going plastic deformation in compression? Will inputting the experimental data coming from the tensile test be enough for the solver, or do I need to add the data specifically from the compression test?
      Plus, do I need to add the Tensile/Compressive Yield and Ultimate strengths as well before or after adding the tabular data for plastic strain (under multilinear hardening)? If not, then if I do so, will it create a problem for the solver?
      I would be extremely grateful if you can answer these queries. Thank you :)
    • Rameez_ul_Haq
      Subscriber
      ,if you could answer this, that would extremely helpful :)
    • Erik Kostson
      Ansys Employee
      Hello

      I would recommend our course on metal plasticity which we offer for free:
      And also all other recommended courses there (so deviatoric stress, linear material, getting started with mechanical, solid mech.,....).

      Finally the Tensile/Compressive Yield and Ultimate strengths, are used to calculate the factor of safety, and are not used by the solver when doing a material nonlinear analysis. For nonlinear material analysis you need to define a bilinear isotropic hardening curve, or a multilinear....., so a something under the plasticity data /tree in engineering data in workbench. So just as discussed in the above course.

      So go through the course and practice on some simple models as shown there, and then should you have any specific questions, open up a new post as this post is long and has been addressed in regards to when should I use material nonlinearity.
      Hope this help you proceed.
      All the best of luck.
      Erik

    • Rameez_ul_Haq
      Subscriber
      thank you for providing with the full course. I will definitely go through that :)
      If you can just also provide a short answer to this question (which I had asked in mt previous comment), that would do wonders :) Thank you
      "In this video, the data for the tensile strength test is inputted. But what if I have a structure which is going plastic deformation in compression? Will inputting the experimental data coming from the tensile test be enough for the solver, or do I need to add the data specifically from the compression test?"
    • Rameez_ul_Haq
      Subscriber
      ,a short answer from your side will be so much appreciated :)
    • Rameez_ul_Haq
      Subscriber
      ,if you provide your views on the last thing I asked, I would be grateful :)
Viewing 15 reply threads
  • You must be logged in to reply to this topic.