7 Replies Latest reply on Feb 24, 2014 4:16 PM by Jerry Steiger

    Aircraft wing help

    Chester Tate
    Chester Tate Jan 14, 2014 1:12 PM

      I found this cool pic of a cross section of an aircraft wing at http://www.sandwichpanels.org/articles/wing_repair.html.

       

      aircraft wing.jpg

       

      And the materials they used to build it up...It's an article on repairing an FRP aircraft wing...

       

      aircraft wing repair.jpg

       

      ...and i thought it would be cool to model.

       

      To make things simpler for me, I would use a lightweight rigid foam core instead of the Nomex Honeycomb. Then add the layers of FPR (carbon fiber, Kevlar fiber, & glass fiber) to the core's surfaces. I would like to be able to change the size of the wing with some ease, if possible. It would be neat to look at a really big wing. The core thickness would need to increase and decrease with some ease too, if possible.

       

      There are probably a couple ways to model this. I'm looking for an opinion on the best way to approach this. I'm not concerned with drawing the geometry, just modeling it efficiently. For instance, using a solid body for the foam and surface/shell as FRP? One multi-body part? A bunch or parts into an assembly? Any insight would be great.

        • Re: Aircraft wing help
          Jared Conway
          Jared Conway Jan 15, 2014 11:54 AM (in response to Chester Tate)

          start simple

          use a shell to idealize the whole stack up

          if you have sim prem, you can use the composites module

          if not, you'll have to figure out the equivalent properties manually

           

          then from there, get a better understanding of the assumptions made in that stack up method and start deciding if you want to go with multiple parts and mixed meshing.

            • Re: Aircraft wing help
              Chester Tate
              Chester Tate Feb 22, 2014 9:55 PM (in response to Jared Conway)

              Thanks for your reply. Been working on finalizing the concept. brought up a couple questions.

               

              The FPR composite lamainate skin surrounds the foam core. it's linear elastic - isotropic to failure. no problem.

               

              The foam is a rigid polyurethane with a density of 20 pcf. it's nonlinear elastic to failure...i believe, so the problem in becomes nonlinear.

               

              My question is, when defining the custom material for rigid PU foam, is it better to input a SS curve or use the whole slew of test data obtained from the manufacturer? i have all the data i need for to input for a linear elastic - orthotropic material (for example). i'm just not sure it's going to be accurate, or produce accurate results because of the nonlinear SS curve. if the SS curve is best, any idea on how to get it... barring experimental testing?

               

              Cheers.

                • Re: Aircraft wing help
                  Jared Conway
                  Jared Conway Feb 23, 2014 7:35 AM (in response to Chester Tate)

                  What nonlinear material model are you going to use?

                   

                  For accuracy, this is where small tests are best. For example, concerned about the material, build a test, compare to simulation, adjust.

                   

                  In the end, simulation is a compromise. Input what you have, write down assumptions for whatever you idealize and understand how those assumptions will affect your interpretation.

                    • Re: Aircraft wing help
                      Chester Tate
                      Chester Tate Feb 23, 2014 6:57 PM (in response to Jared Conway)

                      <<What nonlinear material model are you going to use?>>

                       

                      Well, i'm concerned about modeling the rigid PU foam accurately. I am assuming this is a nonlinear elastic material. That would be the material model I select. The stress-strain response is nolinear at larger strains. But there are three potential SS curves: tension, compression, and shear. Which is correct?

                        • Re: Aircraft wing help
                          Jared Conway
                          Jared Conway Feb 23, 2014 8:42 PM (in response to Chester Tate)

                          might be worth browsing through the help.

                           

                          the foam sounds more like it is better handled by hyperelastics

                           

                          http://help.solidworks.com/2014/English/SolidWorks/cworks/c_Using_Test_Data_for_Mooney_and_Ogden_Material_Models.htm#sra1317223008841

                           

                          in the end, it all comes down to the assumptions you are ok with. my guess is linear elastic is a good assumption for some cases and lets you get started quickly. with the others you're talking about spending time gathering materia properties..etc.

                            • Re: Aircraft wing help
                              Chester Tate
                              Chester Tate Feb 24, 2014 12:00 AM (in response to Jared Conway)

                              Thanks for your input, much appreciated.

                               

                              I think rigid PU foam is closer to linear elastic. If i limit the strains to 0.10, it keeps things in a fairly linear region.

                               

                              I would have to guess that greater accuracy would come by getting SS curves from test samples and using the nonlinear elastic material model (w/ the ss curve(s)). But I will go with the linear elastic orthotropic option for now. I have the data for that.

                               

                              Thanks for your time. Cheers.

                                • Re: Aircraft wing help
                                  Jerry Steiger
                                  Jerry Steiger Feb 24, 2014 4:16 PM (in response to Chester Tate)

                                  Chester,

                                   

                                  If you can stay in the quasi-linear regime for the foam you will make your life much easier. Hyperelastic material modeling is a royal pain. For rubbery materials you really need to have special parts made and tested by an experienced lab. You typically need all three curves, tension, compression (usually done by a bi-axial tension instead), and shear. I suspect the foam may be a little easier, but I would talk to someone at an experienced lab about it.

                                   

                                  Jerry S.