8 Replies Latest reply on Apr 2, 2013 1:07 PM by Jared Conway

    PLS HELP !!!! SOLIDWORKS FLOW SIMULATION VS AUTODESK SIMULATION CFD

    Addy Smith

      I am working on my undergraduate major research project "Aerodynamic analysis of heavy vehicle".While i am using Solid works flow simulation then i m getting 1648 N (if global goal is selected) or 1548 N (if surface goal is selected while i am selecting all the surfaces of vehicle). Which one should i choose for my result/analysis??

      My second problem is that while i am using AUTODESK simulation cfd then i m getting force of 1948 N, since i m new to CFD, hence i am unable to understand which one is right?? can anyone help me or point out my mistake...i will b very very thank full to you...

       

      Or pls tell me how to trust on my result??

        • Re: PLS HELP !!!! SOLIDWORKS FLOW SIMULATION VS AUTODESK SIMULATION CFD
          Jared Conway

          addy, let's break this into a few main questions:

          1. global vs surface goal, if your vehicle is the only thing in the model, they should be the same assuming you've chosen all of the outside faces of the body. do a quick check with a cube to prove this to yourself. they might not be exact but they should be pretty close. be careful about chosen open bodies or double counting by creating multiple surface parameters and adding them together. there are a bunch of good articles on this in the SWX KB. you can also do some checking with a surface parameter to confirm your surface goal.

           

          2. autodesk vs solidworks simulation? if the setups are the same, i'd expect them to be similar. i'd say what you have is pretty close. assuming your problems are setup exactly the same, i'd say going with any of your values would be reasonable. if there is a difference, it might be because you aren't using the right method to output the value rather than the software outputting different values. the only way to tell if one is more accurate than another is to go out and test it. see 3 for more on this.

           

          3. what value do you report? thats a loaded question. what are you expecting the answer to be? if your expected value is around 1700N, they all seem reasonable to me. if you answer needs to be absolute, you probably can't use CFD, you'll need to do a wind tunnel test. for the purpose of your report you should probably check a couple of known solutions to wrap your head around the fact that there will be a difference between physical and calculated but that it will be close and then all you can do is do your due diligence on mesh improvements (look at solution adaptive) and variance (size of comp domain..etc). if you know something about turbulence parameters you could play with that too but i wouldn't touch it unless you have data. in the end, the idea for these tools it to look at relative performance between designs, rather than absolutes so i think any of your results are valid. if you want absolutes, go out and test it now that you know you're in the ballpark and then come back to the sim later.

            • Re: PLS HELP !!!! SOLIDWORKS FLOW SIMULATION VS AUTODESK SIMULATION CFD
              Addy Smith

              Sir yesterday once again i tried to analyze a truck body via both softwares..and again i get very close result by solidworks flow (it gives Cd = .85 while experimental data was  .86) while in same condition autodesk cfd gave a result Cd=1.15, it can b error of turbulance model??

               

              I jst want to go with solidworks flow becoz im very happy from its accuracy & speed...jst please tell me or provide me best cfd setting (such as mesh sizing, adaptive boundry layer techniques) for MOST accurate calculation of Co efficient of drag...

               

              Thanking You A Lot!!!!

                • Re: PLS HELP !!!! SOLIDWORKS FLOW SIMULATION VS AUTODESK SIMULATION CFD
                  Jared Conway

                  simulation vs autodesk, don't think I can answer that question. my guess would be that turbulence model might be different like you said, in sufficient mesh refinement, input error or output process error.

                   

                  regarding simulation, if your answer matches physical testing, it sounds like your setup, including mesh is ok. but for the purpose of doing your due diligence, there is no right answer to your question. start with mesh level 3 and a reasonably sized comp domain (See the solidworks KB for suggestions, something like 3 lengths in front and 5 to the back and 2 to the top and sides is a good starting point) and then add local mesh to the area of interest or an improved global mesh and see if the results change significantly. if they do, then you need to improve the mesh and/or increase the size of the comp domain. this will obviously take longer to solve, so eventually you will have to make the call on the diminishing improvements in "accuracy" vs speed. the solution adaptive technology can help here by only adding cells in the area of high gradients. It starts at result resolution 6.

                    • Re: PLS HELP !!!! SOLIDWORKS FLOW SIMULATION VS AUTODESK SIMULATION CFD
                      Addy Smith

                      I am using result resolution level 8 (maximum), I jst want to ask should i change minimum gap refinement manually??? and minimum geometry wall thickness manually?? i am unable to undersatnd how to use adaptve mess technology?? i m very greatful if u can tell me in details..pleaseee,,,

                      Wht should b value of following things in mesh setting :

                      1.Narrow channel refinement level

                      2. Refine all cells level

                      3. Solid feature refinement level and other curvature refinement level

                       

                      Should i increase no of cells even after level 8 resolution?? pls reply...

                      thanking you!!!!

                        • Re: PLS HELP !!!! SOLIDWORKS FLOW SIMULATION VS AUTODESK SIMULATION CFD
                          Jared Conway

                          addy, there is no right answer and it would be impossible to answer without seeing the model and testing. what i'd recommend is going through the meshing tutorial in the flow simulation tutorials and going through the technical reference and solving engineering problems documents so that you can better understand how the controls work and what they do.

                           

                          generally we recommend to our customers start with level 3, go to level 5 with a min wall or min gap whichever is smallest and then add additional controls based on what you learn in the tutorials and then go to solution adaptive.

                           

                          for solution adaptive, it is defined in the help and tutorials but the easiest way to turn it on is to go into the calc control options and on the bottom right click to reset it, reset all and move the slider to level 6 or higher. this will envoke solution adaptive meshing for your problem.