15 Replies Latest reply on Nov 14, 2014 7:16 PM by Jared Conway

    What published resources for Simulations using Beam Elements would you recommend?

    Tim Rosenberger

      I have done online searches, help file searches, spent some time in the Forums, and looked back through my Simulation training manuals (Simulation, Simulation Professional, & Non-linear), but have not found a comprehensive treatment of beam elements and their peculiarities.

       

      What other resources have people found related to beam elements?

        • Re: What published resources for Simulations using Beam Elements would you recommend?
          Shaun Densberger

          What type of information are you looking for? Is it beam elements in general, or specific to SolidWorks? Are you looking for general information, or do you want to get into the finer details such as stiffness matrix formulation?

            • Re: What published resources for Simulations using Beam Elements would you recommend?
              Tim Rosenberger

              I have a text which covers general stiffness matrix formulation, but a discussion of how SolidWorks implements that basic formulation is of interest to me, particularly how the SolidWorks implementation treats frame drift (if at all).  Further, a discussion of using beam elements in large displacement and non-linear studies is of interest.

                • Re: What published resources for Simulations using Beam Elements would you recommend?
                  Shaun Densberger

                  You'll most likely not find any information on the specific inter-workings within SolidWorks; that kind of stuff is typically kept internal. I'm not familiar with frame drift (this is a creep effect?), so I can't say whether or not this is something that can be captured in SW. As for a general discussion on nonlinear beam elements, Dr. Felippa at University of Colorado at Boulder provides a good start.

                  http://www.colorado.edu/engineering/cas/courses.d/NFEM.d/NFEM.Ch10.d/NFEM.Ch10.Slides.d/NFEM.Ch10.Slides.pdf

                  • Re: What published resources for Simulations using Beam Elements would you recommend?
                    Jared Conway

                    Geostar/cosmosm manuals are likely your best bet. But a test against known solution is probably better.

                     

                    to answer your question about frame drift and nonlinear beams would need some references or more context

                    • Re: What published resources for Simulations using Beam Elements would you recommend?
                      Tim Rosenberger

                      Frame drift is simply deflection of part of a structure tangential to the direction of its major static load component.  This creates an eccentricity between the nodes where load is applied and the nodes where reactions are taken up by the supports.  This eccentricity creates or amplifies bending moments in members, in a way often referred to as secondary bending moments.

                       

                      An assumption of (very) small deflections might (but does not need to) exclude the effects of frame drift.

                       

                      I would guess that large-deflection formulation would take drift into account, but I'd like like that assumption verified.

                       

                      And the question remains if SolidWorks Simulation has implemented a non-linear formulation of Beam Elements, and if so, to what degree?  What nonlinear options would be available?  I could experiment, but I can't justify taking the time right now to try to work through all the permutations myself - I need to quickly evaluate options for using Simulation to improve on spreadsheet load calculations I am doing now, which are loaded with conservative assumptions to keep them in a closed algebraic form.

                       

                      In some ways my question is an extension of an earlier question (Can SolidWorks simulation perform FEA analysis to building code for structures?) in that I am looking to see how far SWX SIM takes the data to include so-called P-D and P-d effects (two different secondary bending effects), and whether I can just load the beam analysis results into my design check spreadsheet or if I need to add correction factors to account for those effects, as required by structural codes such as AISC 360 or CSA S16.

                        • Re: What published resources for Simulations using Beam Elements would you recommend?
                          Jared Conway

                          are you evaluating whether solidworks simulation will be a fit for your needs? if so, your reseller should be able to help you with benchmarking a problem to give you confidence rather than relying on review of the documentation. IE, the proof is in the pudding.

                           

                          what i'm not clear on is what you mean by implementation of nonlinear with beams. are you talking about large displacements? changing of the stiffness based on load? contact?

                           

                          regarding analysis relative to code, the previous posting from bill was spot on. this is a general purpose FEA tool so it isn't built to automate the setup and running of an analysis particular to a specific code. (well except for pressure vessel but it is somewhat open ended and requires the engineer to plug in the specific values and tests) it seems like to me P-D and P-d effects are both things you have to include in hand calcs whereas simulation is solving the problem directly and should not necessarily need the adjustment. corollary to for example needing to use 2 different equations for calculating beam deflection by hand based on their lengths but that not being necessary in sim because the problem is being solved directly. but at the same time, a reference here would be helpful. maybe your hand calcs, a sample test...etc. based on your first comment about frame drift, to me it means that rather than adding the eccentricity as an adjustment, you would actually have to model that into your analysis.

                           

                          have you seen this done in another tool?

                            • Re: What published resources for Simulations using Beam Elements would you recommend?
                              Tim Rosenberger

                              (I apologize in advance for not addressing your points in order, but there is a logic to my presentation below)

                               

                              I already use simulation for other types of analysis. 

                               

                              I don't have sufficient confidence in a simple benchmark problem to make assumptions about how the program is going to behave in a full, complex problem, and benchmarking a complex problem is largely pointless-the whole point is that my closed-form "hand" calculations (actually Excel) contain too many simplifying assumptions, leading (most likely but not perfectly guaranteed) to solutions that are very conservative, overly costly.

                               

                              I have seen programs such as RISA 3D that are used by engineers that only do structural work, but with SolidWorks in hand, I am finding it difficult to justify to my company managers the cost of another software package if SolidWorks is capable of doing the work.

                               

                              I should know one way or the other what Simulation is doing, so that I am not either failing to include the desired "secondary" effects, or including it twice and judging a perfectly good structure to be unstable.

                               

                              Let me try some of my basic questions over again, restated as concretely as possible.

                               

                              1) in linear static analyses with beam elements, (without large displacements turned on) does the solver iterate on the displacements, moving the load application points to their displaced locations and recalculating the deflections based on changed moments or does it just calculate displacements consistent with the initial loading geometry and stop?  Does it matter which solver is used?

                               

                              2) Can a linear static analysis with beam elements be done with large displacements turned on?  What is then done differently?  Again, does it matter which solver is used?

                               

                              3) I'll leave non-linear beam elements alone for now, but one of the areas of interest I would have is elasto/plastic behavior.

                                • Re: What published resources for Simulations using Beam Elements would you recommend?
                                  Shaun Densberger

                                  1) in linear static analyses with beam elements, (without large displacements turned on) does the solver iterate on the displacements, moving the load application points to their displaced locations and recalculating the deflections based on changed moments or does it just calculate displacements consistent with the initial loading geometry and stop?  Does it matter which solver is used?

                                   

                                  It does the former; re-solving the problem based on the deformed shape and stresses is non-linear in nature.

                                   

                                  2) Can a linear static analysis with beam elements be done with large displacements turned on?  What is then done differently?  Again, does it matter which solver is used?

                                   

                                  Large-displacement is be definition a non-linear event, so no.

                                   

                                  3) I'll leave non-linear beam elements alone for now, but one of the areas of interest I would have is elasto/plastic behavior.

                                   

                                  What specifically? Can you capture plasticity with beam elements?

                                  • Re: What published resources for Simulations using Beam Elements would you recommend?
                                    Jared Conway

                                    you seem to have a high level of concern about the accuracy of the solution and what solidworks simulation will do

                                    if that is the case, i highly recommend running a benchmark analysis

                                    i think all your points about what the software can and can't do have been covered but to be certain it will work for your needs, i wouldn't skip a check

                                    personal opinion, but highly recommended based on experience

                                • Re: What published resources for Simulations using Beam Elements would you recommend?
                                  Shaun Densberger

                                  Frame drift is simply deflection of part of a structure tangential to the direction of its major static load component.  This creates an eccentricity between the nodes where load is applied and the nodes where reactions are taken up by the supports.  This eccentricity creates or amplifies bending moments in members, in a way often referred to as secondary bending moments.

                                   

                                  OK, so we're talking about buckling here, correct?

                                   

                                  An assumption of (very) small deflections might (but does not need to) exclude the effects of frame drift.

                                   

                                  It depends on how large the frame drift is. If magnitude of the drift is small enough, you could do a linear buckling analysis, however, I never feel comfortable taking the values from a linear buckling analysis and applying any serious weight behind them given the inherent non-conservative nature of the analysis.

                                   

                                  And the question remains if SolidWorks Simulation has implemented a non-linear formulation of Beam Elements, and if so, to what degree?  What nonlinear options would be available?  I could experiment, but I can't justify taking the time right now to try to work through all the permutations myself - I need to quickly evaluate options for using Simulation to improve on spreadsheet load calculations I am doing now, which are loaded with conservative assumptions to keep them in a closed algebraic form.

                                   

                                  SW does support NL formulation for beam elements, but it does not account for large displacements. Since large displacement formulation is a core requirement when doing a non-linear buckling analysis, it sounds like you'll need to look at a different FE code.