10 Replies Latest reply on Sep 4, 2014 4:23 PM by Jared Conway

    having problem with a point load.

    Max Aguirre

      Im trying to do a dynamic analysis on a cylinder structure. As i need to excite the structure, im trying to put a point load at the end of the structure (its a traffic light). I use the split line function to make a small circle and to put the load on it to simulate a point load. But, when i run the analysis, this load does not produce anything, neither stress or displacement. Do you know what im doing wrong? thanks!

        • Re: having problem with a point load.
          Jerry Steiger

          Max,

           

          You will probably have to post your analysis, as your description says that you are doing it the correct way. If you don't want to attach your analysis, some screen dumps might help someone figure out what is going wrong.

           

          Jerry S.

            • Re: having problem with a point load.
              Max Aguirre

              First, thanks for taking the time to help me.

               

              Sure im gonna attach the file but now its weird, in the file that i attached, now the load is making some stresses and deflections, im happy with that but i have another problem. The purpose of this study is to study the vibrations on a traffic light, therefore, im doing dynamic analysis. Right now, im trying to determine a free vibration of the structure, therefore, as you will se in the study, there is anything else than the structure and a load and i have put some sensors, at the middle and at the end of the mast arm, to make a response graph of acceleration vs time. But im not having the results that i supposed to have. The response graph does not look how it should be.

               

              My response graph looks like this:

               

              acceleration_y.jpg

               

              and it should look something like this:

               

              Screen Shot 2014-08-31 at 3.03.15 PM.png

               

              so as you can see, there are different.

               

              It is my first time using solid works so most probably im doing some mistakes.

               

              Also, another question is that when you put the damping of the structure, it should be in percentage (0.2% in my case) or as a decimal?

               

              thanks!

                • Re: having problem with a point load.
                  Jared Conway

                  where are you getting your loading from? you kind of have a load that is 1000lb that is there for 2s and then released at 3s but is that what you're looking for?

                   

                  damping 0,2 percent would go in as 0.2/100=0.002, but you've asked for damping on 15 frequencies but are only including 5

                   

                  and are you sure that second plot is what you expect? i see normalized acceleration vs acceleration, and i don't think the damping is equivalent. in sim we're talking modal damping.

                   

                  i'd recommend going through the tutorials and through the sim prem dynamic training first. if you're looking for online training or something specific to your application, check out the training section of our website at www.hawkridgesys.com or contact me direct, jared@hawkridgesys.com

                  1 person found this helpful
                  • Re: having problem with a point load.
                    Shaun Densberger

                    How did you obtain the second graph? Physical test? It looks like your FEA model doesn't have the traffic light represented in it (mass point or idealized geometry); was this included in whatever method you use to generate the second graph?

                • Re: having problem with a point load.
                  Max Aguirre

                  Thanks for the replies!

                   

                  Well, i have the load only for 2 seconds because the purpose of the load is just to produce some vibrations on the mast-arm, and then graph the acceleration vs time to see how the damping of the structure is reducing the acceleration through time. The magnitude of load is not matter as i just need to excite the structure. However, your replies make me think that the magnitude of the load would affect the number of frequencies that the structure would have, im not sure. As im trying to analyze vibrations due wind forces, i just need to excite the first frequency of the structure, which when i do a modal analysis, it is close to 1.86 hertz. So, reading your comments i would change to 1 frequency both in the properties of the analysis and in the modal damping.

                   

                  Also, i got the second lab from a research paper and what they did was a physical test but the purpose of this simulation of solidwors is to simulate that, so thats why i need a graph that behaves similar than that!

                   

                  I really appreciate your ideas!

                    • Re: having problem with a point load.
                      Shaun Densberger

                      "However, your replies make me think that the magnitude of the load would affect the number of frequencies that the structure would have, im not sure."

                       

                      The magnitude of the load has no effect on the what frequencies the structure has. All 'real-world' structures have an infinite number of modes of vibration (if we assume they are truly continuous). All finite element models will have a finite number of modes of vibration, where the highest mode of vibration is typically a dilatation mode of one of the elements. The modes of a vibration for a finite element model are determined by solving an eigenvalue problem with the model's mass and stiffness matrix. This means that the modes of vibration associated with a structure are purely a function of the stiffness of the structure, and the amount and distribution of mass within the structure (if a linear assumption holds true, which for most cases it does).

                       

                      Now, as far as the load in concerned, the magnitude has no effect on what frequencies are excited; however, the temporal shape of the load does. In other words, how your load varies with time will determine what frequencies are excited. If your load is constant with respect to time, then the 'period of vibration' for this load is infinite; this means that the frequency of the load (1/T) is 0 Hz (ie no motion) and zero modes are excited. If you load looks like a direct-delta function (ie a impulse load over an infinitely small window of time), then the frequency of your load is infinite meaning that all frequencies are excited (this is partly why shock loads can be so damaging).

                       

                      "i just need to excite the first frequency of the structure, which when i do a modal analysis, it is close to 1.86 hertz. So, reading your comments i would change to 1 frequency both in the properties of the analysis and in the modal damping."

                       

                      Exciting only one frequency is typically not enough, and this will result in modal truncation error. Dynamic analyses are rather tricky because you need to make sure you capture enough modes to reflect the excitation nature of your forcing function. You also need to make sure that you have good modal results, as they are the foundation of a good dynamic analysis.

                       

                      "Also, i got the second lab from a research paper and what they did was a physical test but the purpose of this simulation of solidwors is to simulate that, so thats why i need a graph that behaves similar than that!"

                       

                      So this paper measured the time-history response of a traffic light post to some dynamic load? Can you provide a link to the paper? Are you sure you've accurately reproduce what they tested in your FEM?

                      1 person found this helpful
                        • Re: having problem with a point load.
                          Max Aguirre

                          "If your load is constant with respect to time, then the 'period of vibration' for this load is infinite; this means that the frequency of the load (1/T) is 0 Hz (ie no motion) and zero modes are excited"

                           

                          I think this part i have it correctly because, as i understand, the load is applied for only two seconds so that must produce some excitation in the structure.

                           

                          "Exciting only one frequency is typically not enough, and this will result in modal truncation error."

                           

                          I think this answer my question for an error that im getting right now haha! When i changed it to just analyze 1 frequency, it start giving me an error of "no excitation has been found", and must be for what you just told me!

                           

                          So this paper measured the time-history response of a traffic light post to some dynamic load? Can you provide a link to the paper? Are you sure you've accurately reproduce what they tested in your FEM?

                           

                          They make a small displacement at the end of the mast-arm and released it. All the dimensions of the structure, thickness, length and everything are the same or really close. The only thing that would be different a little bit is the type of steel which the paper doesn't specify which one it is.  I attached the paper, i posted the wrong picture last time but as you can see they behave similar. The graph it is on page 17.

                           

                          Im sorry if im doing some amateur mistakes, first time using solid works and nobody is helping me, i really appreciate your help!