13 Replies Latest reply on Feb 5, 2008 10:59 AM by Ian Hogg

    problems with piston linkage (again)

    1-AKNURG Jan 20, 2008 11:35 AM
      Hi i fairly new to motion and i have tried to simulate my linkage in motion using the intelli motion builder but it has not seemed to work. i have used rotary simulation in solidworks and that workd but i would like to do the same in cosmos motion so that i can work out the magnitude of forces generated at the pivot points.

      can anyone help me? a forum member called Ian helped me lasted time but was unsure how he did it.

      thanks

      Notay
      • 2104 Views
      • Categories: Flow Simulation, Simulation
      • Tags:
        • problems with piston linkage (again)
          Ian Hogg
          Ian Hogg Employee Jan 20, 2008 1:30 PM (in response to 1-AKNURG)
          Hi Notay,

          In motion (2007) you have 2 options. You can apply motion to bodies, or to joints. Applying it to bodies is very similar to what you do in Physical simulation etc, the key difference is you need to define a coordinate system to define which direction the motion(s) will act. Apply it to joints is just a matter of right clicking on the joints and editing their properties and clicking on the motion tab to define motion on a specific joint freedom (eg Rotate Z for a revolute joint).

          To some extent, this has been made more intuitive in 2008, where you appy motors or actuators the sames as in physical simulation and SW animator. You can only define a motion in one direction, but it realtes more directly to what you would have in the real world than in the form of what the solver takes the information (constrainted motion in specific degrees of freedom between a pair of markers).

          I thought you had gone through the tutorials, and that should cover the basics of looking at applying motion and looking at results. Training is really the next step to take the best advantage of the tool.

          Cheers,

          Ian Hogg
            • problems with piston linkage (again)
              Lenny P
              Lenny P Jan 20, 2008 6:31 PM (in response to Ian Hogg)
              Ian or Notay,

              Can any of you send me these Motion tutorials you mention. I don't seem to have any in the student edition?

              Thanks.
                • problems with piston linkage (again)
                  Ian Hogg
                  Ian Hogg Employee Jan 22, 2008 3:08 AM (in response to Lenny P)
                  Hi Lenny,
                  Do you have a folder: C:\Program Files\SolidWorks 2007\COSMOS\Motion For SolidWorks\Tutorials?

                  Prior to that there were a couple of motion examples in the samples folder.

                  Cheers,

                  Ian
                    • problems with piston linkage (again)
                      1-AKNURG Jan 22, 2008 11:48 AM (in response to Ian Hogg)
                      Hi Ian, I will have a go at it and then get back to you.
                      thanks

                      Notay
                        • problems with piston linkage (again)
                          1-AKNURG Jan 26, 2008 2:50 PM (in response to 1-AKNURG)
                          Hi Ian, I followed the same method to the way you did it last time. I begun using intelli motion builder to export the moving and ground parts in and then created the required joints. of course the intelli motion builder created its own joints but i suppressed these as these joints were not present in the assembly model you helped me with last time.
                          however when simulating, the linkage breakes and the orientation of the parts change why is this? i have uploaded the up to date version of the assembly for your conveneience. I am using solidworks 2007

                          thanks

                          Notay
                            • problems with piston linkage (again)
                              Ian Hogg
                              Ian Hogg Employee Jan 28, 2008 5:12 PM (in response to 1-AKNURG)
                              Hi Notay,
                              There are serval things causing problem with this model.
                              You were missing a joint defining the rotation of the crank with respect to ground, and this was causing complications. I see you use lots of distance mates to position some geoemtry with respect to one another, but this is not very useful from a mechanical/mechanism aspect.

                              You need to think about trying to keep your assemblies clean and structured. Putting everything at 1 level is going to make your life harder than it needs to be. When working with motion, think of brouping things into sub assemblies based on if they move as one entity. This carries over well into other areas also, and it make mating the system far simpler and easier to follow.

                              I've attached an assembly (2007sp5) with sub assemblies defined and a cleaned up mate system. It is not perfect, but should give you the idea about how best to approach these. Notice that I didin't really have to do anything in COSMOSMotion, the physical simulation data carried over as is and works. There are still some redundancies that can be cleaned up to make the assembly more robust.

                              Cheers,

                              Ian
                                • problems with piston linkage (again)
                                  1-AKNURG Jan 29, 2008 1:10 PM (in response to Ian Hogg)
                                  hi Ian, i couldnt download the file you put on as it was not attached to your reply, please can your provide it.

                                  i thought i was grouping things when i created the joints as motion was grouping entities itself. so what your saying is i need to reassembly the linkage in subassemblies? would this solve the problem. i also need the distance mates because they define the geomtry of the linkage, what would you suggest to be the right way to do it? thanks for all your help to date

                                  thanks

                                  Notay
                                    • problems with piston linkage (again)
                                      Ian Hogg
                                      Ian Hogg Employee Jan 29, 2008 1:22 PM (in response to 1-AKNURG)
                                      Hi Notay,

                                      Not sure why the doc did not attach. Anyway, attaching it again.

                                      If you do things at the assembly level, it makes life easier for all the other areas, so it's a good practice to modularize. I understand the reason for the distance mates, but just be careful you are only controlling axis, not planes (sometime updates can have unexpected consequences). I prefer to use geometry since it makes defninition of the offsets more intuitive. For your crankshaft, you can still define the axis relative to the link mount, but do it by having a bearing for the crankshaft and this will help you with joint definitions.

                                      Cheers,

                                      Ian
                                        • problems with piston linkage (again)
                                          1-AKNURG Jan 31, 2008 5:27 PM (in response to Ian Hogg)
                                          Hi Ian, thanks for your reply

                                          i couldnt determine the forces acting at the pivot points from your model, i tried to add a inline joint to the floating pin at the conrod end to the conrod_part, but this was unscuccessful and solidoworks did not simulate it. is there a way that i can find the forces at the pivot points, i tried using the concentric joints that you added in but i get a straight line as expected as i will not get any forces axially along the pin which i can see.

                                          also i was wondering as you have helped me greatly with my linkage, is it okay if i acknowledge you in my report?

                                          thanks

                                          Notay
                                            • problems with piston linkage (again)
                                              1-AKNURG Jan 31, 2008 7:46 PM (in response to 1-AKNURG)
                                              Hi Ian, i managed to get a force reading on all the pivot points. i just changed the speed to 40500deg/s (6750rpm) and managed to get a max reaction force of 2714N when the piston is a Bottom dead centre which is obvious. one thing that is strange is that the forces are twice as great than the previous linkage that i simulated, that had a max of 1300N at the same speed which is strange.

                                              i applied a plot to the concentric pivot points, was this the right way? i think the reason i was getting 0N as stated in the above post was because the speed of simulation was low.

                                              thanks

                                              Notay
                                                • problems with piston linkage (again)
                                                  Ian Hogg
                                                  Ian Hogg Employee Feb 1, 2008 5:15 PM (in response to 1-AKNURG)
                                                  Hi Notay,
                                                  Plotting the reaction forces for the joints is the correct way to obtain forces. You need to keep in mind what you look at. for example I had just a revolute joint for the crankshaft, so there is a reaction force and reaction moment component. The moment compensates for only having one joint.

                                                  If you were to have a second bearing, then you should end up without any moments and just forces acting on the joints. The reason you don't do this unless you're careful in how you set it up in rigid body dynamics is that redundancies can confuse what the real loads are.

                                                  Typically for a crankshaft I use a spherical and an inline joint at the ends to minimally constrain it, yet get correct bearing loads. If you have more than 2 bearings, then that is statically indeterminate and you really need finite elements to work out the load distribution.

                                                  Obviously if you're comparing this model to a previouis one, check the following aspects:
                                                  gravity term the same?
                                                  rpm the same?
                                                  mass properties the same?
                                                  Linkage geometry the same (length of components similar)

                                                  It's always good to do some simple hand calcs to verify things are set up correctly. If you know the mass of the piston and the acceleration it undegoes, then you should have a good feel for the force in the pin to the conrod.

                                                  Hope this helps you out.

                                                  Cheers,

                                                  Ian
                                                    • problems with piston linkage (again)
                                                      1-AKNURG Feb 5, 2008 9:11 AM (in response to Ian Hogg)
                                                      Hi Ian

                                                      materials have changed, but the big change was with the use of a 2 arm bar where the pivot points are, so this could be the reson why there was twice the force at the pivot point.

                                                      also on the previous linkage there was an in-line joint at the pivot points and on this linkage there is a concentric joint for the pivots and the reaction force was based on these joints, should it matter that the joints are diiferent, would this affect the magnitude of force generated in the simulation?

                                                      thanks

                                                      Notay