10 Replies Latest reply on Feb 17, 2015 2:33 PM by Bill McEachern

    Internal flow into external space

    Plamen Dimov


      as a part of a collage project I'm trying to run a flow simulation for a hydrotechnical structure. The model consist of pipe structure, at the exit of which there is a outer structure (stilling). The point is to examine the water behaviour after the outlet of the pipe. At this point I've modeled the geometry but I'm struggling with the flow analysis. Is there a way to combine internal and external conditions in one analysis, as the main goal is to examine the characteristics of the flow in this kind of situation.

      Any help will be greatly appreciated,

      Thank you

        • Re: Internal flow into external space
          Amit Katz

          Are you trying to model the mixing interface between air and water? As far as I know, this software cannot do that. You can only deal with 1 fluid type (gas or liquid) over the domain.

          • Re: Internal flow into external space
            Jared Conway

            can you draw a diagram using paint or something that shows the flow conditions you are trying to take into account?


            the general answer about internal and external is yes, as long as the fluid type is the same. so for example a pipe flowing from inside some system out into an environmental space can be done by starting with an internal flow going to an external. the analysis type would be external.

              • Re: Internal flow into external space
                Plamen Dimov

                Thank you for the answers,

                yes - the reason I asked generally if this is possible is because I was aware that it's not possible to model interaction between two different phases and frankly I'm not quite sure if this situation applies to that case. In this situation the goal is to visualize the interacion of the flow with the given geometry, after it leaves the internal system.

                The flow conditions in this problem are:

                - Volume flow at the end of the internal system: 90m3/s;

                - There is no dynamic pressure applied to the walls of the pipe.




                Having those conditions in mind - I tried to run an analysis using external flow and separating the pipe element with lids, thus applying those two boundry conditions to both the lids (inlet volume flow set to 90m3/s and outer pressure set to enviromental). In the initial conditions menu in the general settings I leave the values to default (all velocity parameters are set to 0 m/s).

                Running a calculation in these settings leads to situation in which the fluid doesnt leave the internal space and nothing really happengs in the rest of the model.



                  • Re: Internal flow into external space
                    Bill McEachern

                    Your problem is so ill defined and I have no idea what you are tying to get done. Are you trying to get the flow over the structure at the end with a pipe flow of 90 m^3/sec? The statement no dynamic pressure on the pipe is physically impossible if there is fluid motion. You have an internal boundary condition which make no sense either - boundary conditions go at boundaries not in the middle. What do you want know?

                      • Re: Internal flow into external space
                        Plamen Dimov


                        I will try to express myself clearer.

                        The structure at the end is meant to diffuse part of the enegry of the flow after it leaves the tunnel (pipe). Then - part of the diffused flow leaves this system without causing destructive consequences to the structure afterwords (after the model).

                        My goal is to visualize the interaction of the flow with the following structure.


                        Section A-A:


                        Section B-B:


                        Section C-C:





                        The tunnel is a part of a larger system and is designed (with it's section geometry and slope) to contain volume flow of 90m^3/sec, without the flow interacting with its highest point (don't know if I'm expressing it absolutley right in english) - so relatively speaking, there isn't dynamic pressure applied to it.

                        The outer structure (after the tunnel - the internal section in this case) is designed for flow with velocity of ~11m/s and volume flow rate of 90m^3/s.

                        The idea of the model is to examine the flow behaviour and how it matches the calculations done, when designing the outer structure.


                        I'm trying to apply inlet volume flow ot 90m3/s at the tunnel (right part of the model) and examine how the flow behaviour changes in the current system.

                        My question is how to set boundry conditions in this situation (a model which, as I imagine it - analysis type is set to external and also contains internal part - the ending of the tunnel) or if I should take an entirely different approach at solving this (if this is not possible at all). I have already done analysis in SW for different parts of this entire project and achieved sufficient results, but only using internal analysis.

                  • Re: Internal flow into external space
                    Plamen Dimov

                    I achieved some results during the weekend. Any input will be appreciated.

                    Ok, so I created a new model, which contains larger part of the environment - having in mind that the computation domain shouldn't contain external spaces which souldn't be affected by the flow:


                    After that I inserted just one lid at the beginning of the model to define the volume flow rate (it's defined as time dependent function with a maximum of Q=90m^3/s). This procedure actually achieved some results that are similar to the calculations I did on paper for this problem. Some velocity values are similar to those expected.




                      • Re: Internal flow into external space
                        Amit Katz

                        From those illustrations your posted, this does seem to be a problem where there is a liquid/gas boundary. You cannot model this in Flow Sim. What you did was simply extend your pipe over the feature. I'm not sure how accurate your results are going to be, my gut feeling is "not very accurate".

                          • Re: Internal flow into external space
                            Plamen Dimov

                            Thanks for the input Amit.

                            By no means I'm suggesting that this represents 100% the reality of the situation especially since I'm kind of a beginner to SW FS.

                            For the internal part - the correct velocity of ~11m/s is achieved for the given value of the volume flow rate, so at least a tiny bit of the modeling is correct. As for the values for the external part - I guess the accuracy of the results is related to the question how much of an impact would the enviromental air have on the given flow. I understand that Flow Sim can't model interaction between two different phases, but I'm not trying to mix for instance 5m^3/s water with 2m^3/s air. Call me crazy but I don't think water flow with Q=90m^3/s and V=11m/s would meet great resistance from the enviromental air. Surely there are things to be considered such as the aeration of the flow when leaving the system but again - it's about the impact of the values.

                            Since Jared Conway stated that this can be done as long as the fluid type is the same - I tried to model this problem as approximation. Observing the trajectories and the distributions of velocities - things kind of make sense to me. If the results are even roughly correct - that's good enough for this phase of the project.

                        • Re: Internal flow into external space
                          Amit Katz

                          Flow Sim models the conservation of mass, that's why you're getting the average velocity that you expect at the exit - because you set the input flow rate as a boundary condition.

                          The real question here is what you're trying to learn from the simulation. If you're interested in the behavior of the water as it passes your baffle at the outlet of the pipe, then I just don't know how well you can rely on these results. What you have actually modeled is pipe flow that is dumping into a larger pipe (with a sudden expansion), and in that larger pipe you have the baffle (what you refer to as a "stilling".) 

                          You have to ask yourself if you can really expect the water in this model to behave as it would in the actual physical structure. The hydraulic jump, the pressure loss, and the recirculation zone are probably not accurate to what they would be in real life.