6 Replies Latest reply on Mar 1, 2013 12:07 PM by Bill McEachern

    Rudder Flow Analysis

    J. Rev

      Hey all,

       

      I am trying to run a flow simulation on some new rudder designs. I basically want to figure out some lift and drag forces. To do this I was looking at forces in the x and y direction. Attached is the solidworks file. When I run the simulation at a 7 degree angle of attack (4m/s in x direction, 0.49m/s in the y direction) the x component of force (the drag part) is indicating that it is negative, implying that the rudder has negative drag. I double checked and am not mixing up any signs, and it appears to have significant lift (800N) or so. I am a student doing an independent study and trying to learn the software as well, so any help would be greatly appreciated.

       

      A secondary question: I have been struggling out how to make my model a bit more complex. The current simulation just restricts my fluid range to where the water line is (343mm down from the top of the rudder). It would be much better if I could have two fluids, air above that line and water below, as i can experiment with different mounting solutions for the rudder head which connect just above the water line.

       

      Best,

      Josh

        • Re: Rudder Flow Analysis
          Jared Conway

          after running your analysis, i'm going to have to think a bit about the force direction at the angle of attack. my first thought is that it could be because you don't have a "closed surface" in flow which would cause issues with force output (or any integral value). my second thought is mesh density and size of the computational domain. i would try running your rudder as if it is completely submerged and make the comp domain larger and use solution adaptive meshing to see if that helps. when you're outputting the force, make sure you include all of the faces in a single goal or surface parameter.

           

          regarding your second question, that can't be done with flow simulation. you need something that can handle free surface to do that analysis. i can't really think of a good workaround for it unless you put a thin barrier at the surface and consider them separate regions that don't interact. but then you have to ask yourself what you want to learn by learning about what is above the water line.

            • Re: Rudder Flow Analysis
              J. Rev

              Hey Jared, Thank's for getting back to me. I tried all of the things that you suggested. I even unchecked the "ignore internal gaps" etc stuff in case that was causing it to freak out (since there arent any internal voids). I tried it fully submerged with mesh up to 7. same result. I also tried looking at the surface forces instead of just the volume results. all game me the same strange result. I am starting to think that there might be some torque applied by the water such that the innitial force on it is actually forwards, as strange as that may seem. Any more advice you have is greatly appreciated!

               

              Best,

              Josh

                • Re: Rudder Flow Analysis
                  Bill McEachern

                  The problem is that the down stream direction is not aligned with the X direction because of the way you set the AoA. Do the transformation and it should work out. The domain looks quite small for this calss of problem. Further resolving the LE suction peak and the trailing edge base drag will aslo help with accuracy.

                    • Re: Rudder Flow Analysis
                      J. Rev

                      ok, rotated the rudder 7 degrees, and hit it straight on with the flow. The x component of force is now around 70N and the y 250N. It still seems a little high in the y and low in the x compared to experimental, but it is at least in the right direction.

                       

                      Could you go into more detail about the optimization of the leading and trailing edges?

                       

                      Best,


                      JR

                        • Re: Rudder Flow Analysis
                          Bill McEachern

                          If it were me I would use a split line to make the LE area selectable and add a mesh control - pretty fine so that the non cad geometry view looks like a good approximation of the leading edge. I would do the same at the trailing edge. sometime this requires using manual mesh controls. No time to explain that here.

                      • Re: Rudder Flow Analysis
                        Jared Conway

                        hi, sorry, haven't had a chance to take a look at this further. it might help to post your updated model.

                         

                        a couple of suggestions, what are the expected forces? this would be best to compare against what you're getting in the simulation.

                         

                        it might be worth running a flat plate of similar size/shape to see how it compares with your fully featured model.

                         

                        also, you've turned the flow (my guess because of the simulation KB suggestions), you might want to try rotating the model to see what happens with the results. this would just verify your method.