5 Replies Latest reply on Nov 6, 2017 11:29 AM by Dustin Perreault

    5 Stage Single Impeller Analysis

    Dustin Perreault


      I am new to SolidWorks Flow Simulation, however I have attempted to do my research before posting here, and searched the forum so if this topic has been covered I appologize.


      Project Details

      5 Stage Pump

      1000HP Motor

      Inlet Pressure = 130psi (educated guess)

      Max Operating Pressure1440psi

      Fluid = Gasoline & Diesel


      • Specific Gravity = 0.713 @ 60oF
      • Kinematic Viscosity = 0.795Centistokes @ 60oF


      • Specific Gravity = 0.885 @ 60oF
      • Kinematic Viscosity = 4Centistokes @ 100oF


      Flow = 1600GPM/2285 BPH BEP

      Top end Ability Flow = 1750GPM/2500BPH


      I attempted to attach the part file that I am attempting to analyze, however, the file is too large. We received a 5 stage pump and after tearing the pump down we found that the impellers were not the original impellers that were in the pump. I am attempting to calculate the pressure head and the flow rate that the specific impeller can produce. Furthermore, the customer wants to know if the goals can be obtained with using the 1000hp motor.

      My attempt at the flow simulation.

      Custom fluid was used with a Density of 713kg/m^3 and a dynamic viscosity of 6e-006Pa*s

      The inlet and outlet “pipe” was extended greatly because vortices were crossing the boundaries.

      A rotating region was created that engulfed the entire impeller with a 0.1” offset to ensure that the rotating region did not lie directly on the impeller edge. Local Region Sliding was selected and then retested with Local Regions Averaging. For sliding the default time parameters were used.

      A boundary condition was applied to the “outlet pipe” walls that contacted the fluids, these walls were set to stator.

      A “measuring ring” was created in order to measure the bulk average static pressure directly after the fluid leaves the impeller. This component was disabled in the component control settings.


      After researching it was determined that in this specific scenario the boundary conditions must be set as follows:

      Inlet must be flow rate or pressure

      Outlet must be pressure


      In our case, we wish to know the pressure head across the impeller and the output pressure is unknown. Therefore I set the outlet boundary condition to 1000psi and used this as a placeholder so to speak, meaning the pressure drop should remain the same regardless of the value of the output pressure. By doing this is the torque value going to change? Inlet was set to 1800gpm.


      For the second study, an inlet pressure of 130psi was applied to the inlet face of the pipe and an outlet pressure of 392psi was applied to the inner face of the outlet pipe. ((1440psi-130psi)/5)+130psi which would be the theoretical pressure after the first stage impeller. (With maximum operating pressure of 1440 psi)


      Inlet Volume Flow Rate-surface goal-inlet face

      Outlet Volume Flow Rate-surface goal-outlet face

      Inlet Mass Flow Rate-surface goal-inlet face

      Outlet Mass Flow Rate-surface goal-outlet face

      Inlet Average Static Pressure-surface goal-inlet face

      Bulk Average Static Pressure-surface goal-inner face of measuring ring directly outside impeller outlet

      Torque on impeller (Z)-surface goal-every face of the impeller

      Pressure Drop-Equation Goal-{Bulk Average Static Pressure}-{Inlet Average Static Pressure}

      Efficiency-Equation Goal-{Pressure Drop}*{Inlet Volume Flow Rate}/{Angular Velocity in Rad/Sec}/{Torque}

      Pressure Head-Equation Goal-{Pressure Drop}*{Inlet Volume Flow Rate}/372.8/{Torque on Impeller}

      Horse Power-equation goal-{torque on impeller}*8.8507*3560rpm/5252

      Where 8.8507 is used for a conversion factor from n*m to lb*in and 5252 is conversion factor as well


      Global Mesh Level = 6

      Local mesh applied to all of the impeller faces in contact with the fluid. Most sliders were advanced at least halfway. Settings top to bottom: 6,6,5,4,3,4

      Units were in SI and results were converted to English units.

      My flow trajectories were believable, and the cut plots for pressure and what not looked to be located in the correct locations.

      I hate attempted many different setups for the study and have yet to obtain believable results. Generally, I seem to get a negative Pressure error, and when I don’t the torque values are much too high which gives me inaccurate horsepower results. Efficiency is generally not realistic either, (sometimes above 1).I know that the software is capable of doing what I want I must just be setting up the simulation incorrectly. If anyone has any pointers or a better way of getting the customer the results they are looking for that would be greatly appreciated.


      The model is of the impeller only, due to the fact that modelling the pump case is a tedious process that would likely take weeks at best. I set up the study in a similar fashion to the advanced flow simulation tutorial of the centrifugal impeller.


      Here is a link to download the Zip file from my google drive.

      Flow Sim 21024.zip - Google Drive

      Also if Jared Conway is still on this forum and ends up reading this I have also uploaded the zip file directly to your Dropbox, and I apologize for the large file size.


      Thanks in advance,

      Dustin Perreault

        • Re: 5 Stage Single Impeller Analysis
          Dustin Perreault

          One thing I caught this morning is the fact that I set up my Horse Power equation wrong. My conversion factor from N*m to ft*lb is not even close, so I apologize for that. It should be Torque*0.737562149*RPM/5252.


          Also my Equation goal for pressure head is actually {Pressure Drop}/(9.81*713)+{Outlet Velocity}^2/(2*9.81)

          I must have copy and pasted it into the forum wrong.

          Not sure how I screwed that up so bad. This brings my numbers much closer to being realistic, I have refined my mesh further and fixed the equation, and so far there are no warnings. Hopefully, with the refined mesh, I will obtain results that are more realistic.

            • Re: 5 Stage Single Impeller Analysis
              Mark Keown

              I think this looks good.  I found the making the local mesh with a cylinder gave a even velocity plot.

              You have the rotation set at -373 rad/sec or 3561 RPM is this correct?


              The below plot is using the surfaces to create the local mesh.  Using a volume to create the local mesh the plot is radial even.

              Below using volume to create local mesh.

              To make solving faster change the time step to manual and enter a value that is (default time)/10.  Solve with this larger time-step for a while and then change back to the default.  For this sort of problem expect the solve to be > 10 hours.


              Would change Calculation control to All satisfied.


              I am keen to hear what others say.

                • Re: 5 Stage Single Impeller Analysis
                  Dustin Perreault



                  Your Input is greatly appreciated. 3560rpm is correct. Based on the velocity plot that you got I assume changing the local refined mesh to a cylinder is much more appropriate. The Horse Power Results that you obtained seem very reasonable to me. The efficiency value is not quite as high as high as I would have predicted, however that does not mean that it is incorrect. I believe I may have my pressure drop equation wrong still and will look into it. I am going to rerun the study making the changes that you suggested and see what I end up with. I think I am on the right track to solving this problem with your help, so again thank you for your advice.

                  • Re: 5 Stage Single Impeller Analysis
                    Dustin Perreault

                    Just an update. With Marks help, my results improved greatly. My cut plots look much better in my opinion and I obtained similar results to Mark at an outlet pressure of 1000psi. Unfortunately, my boundary conditions are not appropriately set up. I altered the outlet pressure from 1000psi to 750psi (5171000 Pa), in hopes to verify that my initial thought that the results would not change based on the outlet pressure. I was correct in the fact that the pressure head did not change, however, my efficiency and horsepower results varied which is not surprising. Now I am looking into an alternative way to set up my study, if anyone has any ideas it would be greatly appreciated.


                    My results are much closer to what I was hoping for, but that doesn't mean much considering the fact that I am assigning an outlet pressure when in all reality that is what I would like to find out. My dilemma is that an inlet pressure cannot be used in conjunction with an output flow rate, (not suggested in SolidWorks KB, because forcing a uniform flow rate at the outlet is unrealistic).


                    I also cleaned up my equations, (appropriate units).

                    Pressure drop in PSI = 460psi (92*5stages) (still a little low)

                    Efficiency = 60% (still a little low)

                    Power = 1005HP (201hp*5stages) Right at Motor Rating


                    Pressure Contours

                    Pressure Contours Front Plane.png

                    Velocity Contours

                    Velocity Contours.png

                    Pressure Isolines

                    Pressure Isolines.png

                    Velocity Isolines

                    Veloctiy Isolines.png

                    Flow Trajectories

                    Flow Trajectories.png



                • Re: 5 Stage Single Impeller Analysis
                  Dustin Perreault

                  Here is an updated Zip File including my new attempts to solve the problem.

                  Flow Sim 21024 2.0.zip - Google Drive