4 Replies Latest reply on Mar 7, 2016 8:00 AM by Mitch Seymour

    Not sure how to define my flow conditions

    Karlis Banis

      Hello, I am trying to simulate a flow bench, simply because I did not find any other way in flow simulation which could provide me with the same or similar CFM readings as a real flow bench.

      At the moment it is very simple, it is just a pipe with atmospheric pressure opening at one end, a flow orifice in the middle, and the other end is where the problem lives.

      What I need is basically a surface between the atmospheric end and the flow orifice. I need this surface to move volume of air like a fan if you will. And what I need to make sure of is that this "fan" will create a pressure drop of 12 inches of water below atmospheric. I need this exact test pressure at the first section of my bench to proceed with the calculations. This is how a real flow bench would work.

      Then I could simply define the other end as environment pressure and let the air from the fan escape.

      Some of you might say why build a virtual flow bench when I can just calculate the CFM of my engine head directly from flow simulation. If any of you could tell me why am I getting 25 CFM where I should be seeing 200 CFM, please go ahead and tell me, I will abandon the virtual flow bench project right away. And yes, I DO need CFM values similar to a real bench.

      I have attached two pictures, one is my virtual simplified flow bench, and the other one is a CFD analysis of my engine head.


      Over here I have highlighted the piston top where I inserted a surface parameter and picked volume flow rate. It showed me 25 CFM. I have valid information that this head reads close to 200 CFM on a real bench.



      Here is my flow bench. I have highlighted the surface which needs to emulate a fan. If this is not possible, maybe someone can suggest me how else can I define conditions that would create me 12 inches of water pressure difference from the inlet to the highlighted surface and let the air continue down and out the pipe? Thank you.


      Also, I am sorry for this extremely long post but I might have a stupid question. How is it that the mass flow rate is different from one end to the other?

      My surface parameters tell me that there are 0.23 pounds of air entering my bench every second, but only 0.20 pounds exiting it. WHERE does the rest mass go? Or is solidworks flow simulation really inaccurate?


      Thanks to everyone who stopped by to read my problem. I am desperate of this. I really need it to work so please help me.

        • Re: Not sure how to define my flow conditions
          Jared Conway

          is this a commercial, academic or educational project?


          it sounds like you have 2 setups:


          testing the engine head directly, you did not provide any information here about how that was setup. is it pressure/pressure? is it a volume flow rate? generally this setup would be done by forcing a pressure difference in the system and then reading the flow rate or setting a flow rate and pressure and then reading the inlet pressure to determine the pressure drop created by the system


          the flow bench pipe, if it is a fan, why not use a fan? most likely the issues you're having with mass flow rate not matching are either the calculation stopping too soon, boundary conditions aren't valid or insufficient meshing.


          going back to my initial question, if this is a commercial problem, i'd highly recommend working with your var to get some additional support/training to help you get through this. the software is tested and validated so getting good results should not be an issue. it all comes down to the setup. if you're interested in this kind of training we offer it virtually here at hawk ridge, you can get a hold of me at jared@hawkridgesys.com. for academic projects, you could consult your TA or professor for the class, they may have a connection at a VAR. for example we support several universities with these types of projects. and for educational (self education), keep posting here but you may want to take a step back and explain the goal of your analysis and give a detailed model of the flow bench and what testing information you have and what hand calcs you use to validate your designs before cfd.

            • Re: Not sure how to define my flow conditions
              Karlis Banis

              Hey Jared, I greatly appreciate your advice. This is a personal project, so far any educational instances have not been involved, though I am myself a student of mechanical engineering and I do this simply because I see it as my future profession. I will attempt to explain my setup a bit more..

              The model of the cylinder head I have here is scanned using FARO laser scanner. It was then reverse engineered in solidworks over a large period of time because it is not easy. I have information available from a reliable shop in USA that the particular model year for that cylinder head flows approximately 200 CFM on a flowbench by creating a 12" H2O pressure drop from intake to the cylinder. Therefore I directly tried to replicate it. I defined the inner surface of my model airbox as a boundary condition  - environment pressure opening. On my piston top surface I defined a boundary condition - static pressure, which is exactly 12" H2O less than environment pressure.

              That is it, no other boundary conditions were made. I then ran the simulation and after the calculation I inserted a surface parameter, picked my piston top, and marked Volume Flow Rate to be shown. It then showed me values that are abnormal - 24.5 CFM. This was when I started to doubt the accuracy of my setup and tried to build a virtual flowbench to emulate its operation in order to obtain the correct data. I found some basic calculations on PTS flowbench forum that state, when using 2.035 inch orifice in a 10 inch pipe with a pressure difference across the orifice of 36 inches H2O, the CFM going through the orifice is over 300. However, my solidworks model shows me 200, wich a lot closer than before, but still not enough accuracy. Here is the link to the calculations: http://tractorsport.com/download/Orificeflowinapipe.xls

              My model of this test is visible in the picture I already posted. The left end - boundary condition environment pressure, the right end - boundary condition static pressure 13.3949205 PSI.

              It is basically a 10" pipe with a 2.035" orifice.

              I believe I am doing something wrong, but can anyone point out to me what should be done different? I created the study using the flow simulation wizard, everything was left stock, except for surface roughness defined as 50 microinches, air added as the test gas, and the option for advanced narrow channel refinment option selected. I am using SolidWorks Premium 2014.

              Jared, you recommended using a fan. Is there any instruction on internet how to insert a fan in my simulation? I can only find info on how to prepare fan curves... I have no idea what those are for. Thank you!

                • Re: Not sure how to define my flow conditions
                  Jared Conway

                  the fan was only for the simulation where you said you wanted to create fan. if you input flow rate and environmental pressure to simulate a fan, it won't actually act like a fan.


                  I would recommend on concentrating on the model of your cylinder head only


                  first thing, remember that environmental pressure is total pressure at an inlet and static pressure at an outlet. so if you are looking for 12in of pressure drop, is that static or total and have you measured what flow simulation is providing based on goals at the inlet and the outlet?


                  next, remember that these are very rigid boundary conditions. if you read through the kb articles you will see that it is always recommended to move boundary conditions away from your area of interest. this is generally done by adding length to inlets and outlets. 3-5 diameters and even more based on testing.


                  lastly, if you are trying to match a physical test, make sure they are the same. you mentioned something about the manifold, is that included in your setup? do you have the exact fluid conditions..etc?


                  I just worked with a customer that did almost exactly this and got great results between cfd programs and to physical.

              • Re: Not sure how to define my flow conditions
                Mitch Seymour

                Hi Karlis, Im doing a project much like yours for self education and university. I have not managed to get any figures that are even close to what the actual flow bench is telling me. Any chance you could email me so i can ask a few detailed question on set up etc? mitch.j.seymour@gmail.com