2 Replies Latest reply on Aug 23, 2009 6:24 AM by Derek Bishop

    begining to learn simulation

      I have done a coures af 3 days at Solidworks abour simulation, thats all i have got training about fea, when i did my mechanical engineering ages ago there was no such consept of fea. I have done most of the exercises and case studies of the manual supplied by solidworks, but that is not enougn to be good at it.
      This is where i stand.
      Can any body suggest me a way where i can learn more about simulation, i feel i need to learn fea through simulation and not through other programmes.
      can some one please help me.
        • Re: begining to learn simulation
          Gamal Reda
          you may need to learn more about the finite element method itself, I recommend you to read some books about it like "Finite Element Modeling For Stress Analysis, Robert D. Cook" or, " Practical Stress Analysis With Finite Elements"
          • Re: begining to learn simulation
            Derek Bishop

            My suggestion is to start using the software on problems similar to those you are likely to encounter. Run some simple cases and check your results against manual calculations. That will help to build your confidence. I wouldn't get too bogged down trying to understand FEA theory in much greater detail than is given in the training manuals. You don't need to understand how an aeroplane flies to travel on one. Develop  an appreciation for the limitations of the software. Some of the less used areas are full of bugs. How far you want to persevere with these areas is something only you can decide.

            Meshing and defining contacts has been one of the most fustrating and time consuming problems for me. Normally more of an issue where you have mixed meshes and assemblies. Learn to recognise a compatible and incompatible meshes. Learn how to force compatible meshing by splitting surfaces or faces. Learn to identify when you need to define a contact.


            Defining restraints is another big and important area. Simulation has a number of different ways of doing this and it takes a while to get you head around them. For example, take a simply supported beam pinned at one end and supported on a roller at the other end. Every engineer should have encountered this problem. Try modelling that in Simulation with a solid beam. Not as easy as it might first seem but it is possible.


            You need to  find ways to simplify a model and reduce the time to run a study. In a recent case I was able to reduce the time to run an FEA  from 2 hrs to a few minutes by simplifying the model without significantly reducing the accuracy of the model. How many dof's can your hardware handle? How can you reduce the dof's wothout sacrificing the accuracy required? What kind of contacts should you use? Non penetration contacts are very time consuming and I generally try to avoid them but in some cases they are needed. I think you get an appreciation fo these things by using the software.