11 Replies Latest reply on Apr 16, 2010 9:33 AM by Ahmad Zulker

    Steelwork structural design

    Derek Bishop

      I knocked this model up and Simulation study to check out how easy it would be to do a preliminary design check on a steelwork building. The results were pleasing.

       

      Here is the model. Sheet was just flat plate and used to easily apply wind loads. Building is 8 m high, 14 m wide and 63 m long (9 bays, 7m per bay).

       

      Building 1.JPG

       

      The mesh.

       

      Building 2.JPG

       

      And the beam stress analysis.

       

      Building 3.JPG

       

      I thought this was pretty cool and done in a very short time. This kind of thing could be used by mechanical engineers for preliminary dselection of beam size. In my state structural engineers have to sign off any such design.

       

      I heard the automatic definition of beam to shell bonded contact worked well and it did in this case. Had to fudge a few little things in the design to get the nodes to sit in the right place but that was relatively easy and understandable. Seems to have a lot of potential for this kind of thing.

       

      Interested to hear if others use it for this application and what you think.

        • Re: Steelwork structural design
          David Paulson

          Most structural engineers deal only with standard structural members, H beams, C channel, Angle, etc.  Simulation is very powerful in that it allows you to design using thin members, like sheet metal panels.  One of my favorite back burner projects is designing barns fabricated from structural sheet metal panels that are easily bolted together.  In this maner the building structure and the building skin are one in the same, and construction is a fraction of the time and cost.  There are others such as two sheets of plywood bonded to an inner core of polystyrene insulation, Which has a name that I don't recall at this moment.  Or Insulated Concrete Forms wherein the insulating foan is also the form for the concrete structural members.

           

          But the best application for Simulation is the integration of structure and shell.  Thin member analysis almost has to be FEA.

          • Re: Steelwork structural design
            Bill McEachern
            Here is another example:hoisting loads and buckling margins for a packaged building. A couple hours all in from a clean sheet.
              • Re: Steelwork structural design
                Derek Bishop

                Nice Bill. So you checked out the loads with some kind of machine in the building?

                 

                I note you have hidden the skin. I found the displacements of beams were difficult to visualise with the skin. I was able to get around this by hiding the skin. Was this the reason?

                  • Re: Steelwork structural design
                    Bill McEachern
                    I ended up putting the skins in a separate part in order to better visualize the beams. It would be nicer if the hide bodies option worked in simulation - not that big a deal though once you figure out the "rules". The purpose of the model was to check out how well things worked so I put in the gear - motor generator set and made the generator a remote mass and left in the motor to be meshed as solid. I also checked out the remote load to make sure it worked as well. It all seemed to work pretty well.
                      • Re: Steelwork structural design
                        Derek Bishop

                        I have been doing a bit of work checking out the variation in rebuild times with bigger models with a few hundred bodies or more. I found a huge variation in built times based especially related to the use of trims. In broad terms I found it best to avoid the use of trims and substitute this with patterns or move copy commands where possible. This reduced my rebuild times by up to a factor about 5 times. I found the automatic trims used when you build a weldment with one feature and several groups are about as costly in terms of rebuild times as using the trim feature.

                         

                        Have you found similar results. Any draw backs in using the patterned or copies bodies that you've found.

                          • Re: Steelwork structural design
                            Bill McEachern
                            I don't trim any of the structural members as it has no impact on the analysis if beams are used. It would be the last thing I do before heading to drawings. If you are referring to surface trims, if things get slow I do a parasolid export and use that but that has not been required in recent memory. The new enhanced split line has issues with patterning and in the use of multiple contours. At least once when I used a delete face on a patterned split line it would delete the wrong face. I ran away to the wrap feature using the scribe option to fix it.
                              • Re: Steelwork structural design
                                Derek Bishop
                                Yes, good points. I was referring more to trimming structural members. Best to do the analysis before the trimming.
                                • Re: Steelwork structural design
                                  Derek Bishop

                                  This one tested things out a bit more.

                                   

                                  Structure 1.JPG

                                   

                                  Structure 2.JPG

                                   

                                   

                                   

                                  Structure 3.JPG

                                   

                                  I used lots of patterns to set this trial case up. The FEA run took a matter of minutes. Could be used again for preliminary design and member sizing.

                                   

                                  I did have some trouble (using 2010) with some beams being read as solids and not beams. I had to select and convert them to beams. The next problem was that when calculating node positions a number of the beams were not selected and had to be manually selected. I suspect they were the ones that were converted from solids to beams.

                                   

                                  I had to trim the bracing and fiddle around with the locations where bracing finished to get a sensible positioning of the nodes.

                                    • Re: Steelwork structural design
                                      Bill McEachern
                                      I doubt you would have had any beam issues if the whole thing would have been specified using the structural member command. I get the desire to use patterns as it is probably faster but the under lying sketch geometry is used when things are done with the structural member command, hence its ability to reliably get the "nodes" (member connections) as per expectations with out further effort.
                          • Re: Steelwork structural design
                            Ahmad Zulker

                            Hi Derek,

                             

                            Sheet was just flat plate and used to easily apply wind loads.

                             

                            I am curious how you can aplly wind load in SolidWorks Simulation.

                            Can you please tell us more detail step by step on how to apply wind load in SW Simulation ?

                             

                            Thanks