7 Replies Latest reply on Nov 8, 2015 12:06 PM by Thulani Sobekwa

    Non-Linear Material Properties

    James Benting

      Hello!

       

      Has anyone entered non-linear material properties for mild steel?

       

      I have entered standard Young's modulus, also tangent (yielded) Young's modulus of steel.

       

      Any help appreciated,

       

      Thanks

       

      James.

        • Re: Non-Linear Material Properties
          Anthony Botting

          Yes. What in particular are you asking? (sorry, the question is just not clear to me) - Tony

            • Re: Non-Linear Material Properties
              James Benting

              I wanted to check my custom mild steel material properties with someone else's since the non-linear software is new to me. The properties in particular that i wanted to check was the 'tangent modulus) which as i understand is the yielded Young's modulus of mild steel in this case. Also any other settings that differ from the linear analysis.

               

              I have run a non-linear analysis on a product which we have done a physical test on & the results are very different. I also done a linear analysis on the same product which matched with the physical test. This is the reason why i think maybe my material properties are wrong, or maybe something else.

               

              I have created a time curve which apllys the load, then remove's it so i can obtain the maximum & permanent delfection results.

               

              I'va attached a screenshot of the custim material properties i have entered for mild steel,

               

              Thanks,

               

              James.

                • Re: Non-Linear Material Properties
                  Anthony Botting

                  Hi James: I have seen suggestions in training manuals to use something like ETAN is about 10% of E, and then adjust the hardening rules for your test data.  Something like 85% kinematic and 15% isotropic is a good start. You are in a unique position because you have test data, which can be used directly to calibrate your FEA inputs! Keep your results to yourself - you will have valuable data for your material, with consideration of the way it is formed (processed) and how it behaves post-yiield is different if it's rolled, forged, quenched and tempered, etc..: this is information you can sell (as you probably surmised, it's very costly to get test data). I hope that helps! -Tony

                  • Re: Non-Linear Material Properties
                    Jerry Steiger

                    James,

                     

                    I learned about materials in English units, so I translated your values. Young's modulus looks OK, being 29 MPsi, half way between my memories of 28 for SS and 30 for CS. Your yield strength of 31,900 psi may be OK, depending on the type of steel you are using. Not being very knowledgeable about steel, I don't know about your value of 29,000 psi for the tangent modulus is right, but it is only 0.1% of your Young's modulus, not the 10% that Tony suggested. To tell you the truth, his number sounds too high to me, but I'm stuck in a motel room with no access to my material property sources. Wikipedia to the rescue! I would trust Tony's values more than yours. I remember stress-strain curves for steel being very flat, but that is the apparent stress-strain curve. The real stress-strain curve, the one that Simulation wants, is much steeper, since it factors out the necking down that takes place in the test speciman. So it looks to me like your values probably didn't match the test results because you had the wrong tangent modulus, just as you suspected!

                     

                    Jerry Steiger

                     

                    I checked into some more on-line stress-strain curves. It looks like a value of 1% is more likely to be correct, landing neatly between 0.1% and 10%.

                     

                    Message was edited by: Jerry Steiger

                • Re: Non-Linear Material Properties
                  Alessandro Tornincasa

                  Hi James,

                  in some cases I've seen that, by putting the complete Stress-Strain curve, you get much better convergence.

                  Problem is obtaining that data set.

                  • Re: Non-Linear Material Properties
                    Richard Hallman

                    Hi James,

                     

                    I've been using cosmosm for aabout 25 years now & what has worked for me is to use the yield strength and elastic modulus from the Section II Part D of the ASME Boiler and Pressure Vessel Code for the appropriate material and temperature.  For ETAN, I use 2% of the respective elastic modulus.  This is good for moderate load rates & design purposes.  I would use A36 for mild steel.  To be more exact, use test data from the actual material test reports.  I have not done too many of these analyses in Simulation yet, since 2D elements were just added and many of the nonlinear options in cosmosm are not fully implemented yet.

                     

                    For a full true stress-true strain curve, Section VIII Divisdioon 3 of the already referenced Code & ASME FFS-1-Fitness for Service contain equations to generate the full stress strain curve from yield and tensile strengths.  The results are not much different, but you can not use temperature dependent stress strain curves in Simulation or cosmosm.  You can input the yield, ETAN, & modulus as a function of temperature, so you can do very accurate analysis of elastic plastic thermal stress problems.

                     

                    Hope this helps.

                     

                    Richard