21 Replies Latest reply on Jul 16, 2014 7:44 PM by Jared Conway

    Ventilation diffuser design, and flow simulation questions

    Samuel Nilsson



      I have been doing some flow simulations but sometimes they sometimes seem strange and non realistic. Maybe I have done something wrong or the program is i bit wierd. Small changes can make huge differences. Sometines the fluid stops as it hits a surface.


      I also get many error messages like "There is no fluid volume in the project. Please check geometry or boundary conditions." Sometimes I can fix this by increasing the size of the computional domain, making sure there is no small hole or fixing boundary conditions. Sometimes SW is very stubborn and does not cooperate. (At least how I feel).


      I found it easier to do simulation when only doing it on a part. I could not do it on an assembly.


      Another thing I have noticed is that my settings, like preasure, temperature or volume flow doesn´t affect the air flow in the room.


      I am doing a project constructing a ventilation diffuser that is placed on a wall near the celing. I want the air to go across the room near the roof before it sinks down. In many cases the "fresh air" turns to the side towards the wall and drops down to the floor. I wonder why? Any ideas on ventilation diffuser designs (with a connection box) for an airflow that reaches the opposite wall and then spreads out while going down, mixing with the "usead air" and finally returning to an outlet on the same wall as the inlet?





      halvklotsdon med hål.png

      This flow seen above is quite good I think but I want to achive it by using an inlet on a connection box outside of the room. That inlet is facing upwards so the air has to do a 90 °

      turn in the box.




      cfd tilluftsdon 20 maj.png

      This is what the flow usually looks like using a connection box.







      A simle version of a  connection box. I have also tried with a front of perforated sheet.









        • Re: Ventilation diffuser design, and flow simulation questions
          David Paulson

          I would consider your first picture without the connection box to be the flow you would expect in an ideal situation.  The flow pattern looks just as one might expect.  The airflow in such an application will also induce airflow in the room, and I am not sure if Flow is showing this secondary flow in your model.  It would be more helpful if youalso plotted the flow with pipes with arrows which would indicate the directional flow of the air.


          Your design of a connection box is not only converting the air 90 degrees, but it is also increasing the air's velocity at the discharge point.  This type of air delivery is commonly done with a grille with adjustable blades (both horizontal and vertical, typically), which would be another flow model to compare against your connection box design with a slot opening.  Creating a detail view that focuses only the air around and in the connection box may also give you some insight to your design.

            • Re: Ventilation diffuser design, and flow simulation questions
              Samuel Nilsson

              Thanks for your help and comments!


              I am not totally sure what the ideal flow would be but the flow from my first picture seems good.


              Would these adjustable blades be betwen the box and the room or somewhere inside the box?


              Is detail view a separate feature or just by zooming in? I have zoomed in and studied pipes/shperes but it is not easy to see what really happens.


              I still don´t know what is going on with the settings of the inlet flow. Usually I test my ventilation diffuser with 20 l/s. Then I tried it with 3000 l/s to see what happend. It changed a bit and the velocity of the air increased a lot. When I changed back to my "normal" settings the air flow/speed didn´t change. Any Ideas why?


              böj klotdon.png

              By the way, I might bot use a box for a connection. Maybe a 90° bend (125mm) will do better. There would probably be much less swirls. I think I will change the hole pattern of the diffusor seen above and make it a bit flatter. The new hole pattern should compensate for an increased air flow at the top, by using smaller/fewer holes there.


              The ventilation diffuser should preferably have an adjustable k-factor. I am not sure how I would solve that though.

            • Re: Ventilation diffuser design, and flow simulation questions
              Stefan Wozniak

              Wierd things can happen when you have tangent contact betwen faces. In such cases FlowSim has problem to predict where is fluid and where is solid.

              Second thing: in assembly there is a chance that you can by mistake move non fixed components.