Non-reactive and non-diffusive passive scalar transport

    • Ayan Paul


      I am trying to simulate a problem involving convective mass transport on ANSYS Fluent, but so far have been unsuccessful.

      The problem is as follows: An iodine contrast tracer (passive scalar) is carried with blood flow through the 3D mesh of an artery segment. There is 1 inlet and 3 outlets on the artery. The iodine contrast should not chemically react with its surrounding and it should not diffuse into the wall of the artery.

      I believe that the molar concentration for the iodine should change at different spatial zones within fluid domain with respect to time as the iodine contrast, carried by blood, is transported from inlet to the outlet.

      I have attempted to model this problem by following the "Species Transport Without Reaction" guide on ANSYS Fluent documentation.

      I have attached the images of the setup. When I run the simulation on this setup, the residuals for the iodine contrast is not changing (exported data from results also show no change in the molar concentration at different timestep).

      I am using a UDF (pulsatile blood) to define the inlet-velocity and outlet pressure (varying inlet velocity and outlet pressure).

      I am very sure that I am making some mistakes setting up the simulation solution, but cannot identify them.

      Can someone please provide some guidance?

      The screenshots of the solution setup are attached.

    • Rob
      Ansys Employee

      You''ll need to post the images into the thread as staff aren't permitted to follow links and/or download files. 

      Otherwise, if the iodine is added as a volume fraction at the inlet and nothing happens to alter the period when iodine is added (ie a short tracer pulse) then the whole domain will show a constant volume fraction. So, the residual might not change. The critical panel is the inlet, and specifically the species tab. 

      • Ayan Paul

        Hello Rob, thank you for taking the time to answer. I have made some changes taking account of your advice and I would like to ask a few more questions because I am still unsure whether my species transport model is correct.

        My simulation setup is as follows:

        My setup is as follows:

        (i) Pressure-Based solver; transient simulation; 3D artery geometry w/ 1 inlet and 2 outlets; Gravity disabled.

        (ii) Energy off

        (iii) Viscous laminar

        (iv) Species transport with inlet diffusion enabled (more detail about the mixture given in the image below)

        (v) Boundary conditions:

        • no slip condition
        • puslatile inlet velocity for blood
        • pulsatile outlet pressure for blood
        • a rectangular function to define inlet velocity of iodine species
        • wall set to zero diffusive flux

        (vi) Solution Method: Simple

        My quesitons:

        1.  Since the two species do not react, I have chosen the inert-mixture option. The iodine solution will diffuse into the blood (bulk species, fluid domain) and I have selected the "inlet diffusion" option.  How does choosing "inlet diffusion" will affect the simulation?
        2.  I am using a simple rectangular function to define the velocity of the iodine solution at the inlet. The mass fraction of iodine is set to 0.1 at this inlet face. I set an arbitrary value of 0.2 ms-1 for the injection which last for only 1 second. Here is the UDF part of it:

          DEFINE_PROFILE(rectangle_velocity, thread, position)
          face_t f;
          real t = CURRENT_TIME;

          begin_f_loop(f, thread)
          if (t <= 1) // During injection time
          F_PROFILE(f, thread, position) = 0.2;
          else // After injection time
          F_PROFILE(f, thread, position) = 0.0;
          end_f_loop(f, thread)

          I am running the simulation for 32000 timesteps with 0.01 time resolution. Is the above profile correct for the iodine injection?
        3. I used the Gambill method to find out the mixture density and viscosity for varying mass-fraction of each species. I set the viscosity for the for the mixture to the constant amount calculated for iodine mass fraction of 0.1. But I could not set the density to the calculated constant value and so opted for the option "volume-weighted-mixing law" because I had already defined materials properties (density & viscosity) for both iodine and blood earlier. Is this correct given my setup and is there any other way to do it such as writing the value into the UDF and then laoding it from there?

        4. I am struggling to define a inlet for the iodine injection. I selected a face of the inlet to act as the inlet for the iodine solution and the rest of face as an inlet for the blood. This cannot be correct. Both species enter the fluid domain at the same inlet. Because I am defining a seperate velocity profile for iodine injection, is it possible to select a particular coordinate on the inlet face for the second species?

    • Rob
      Ansys Employee

      You've mostly got the setup, but there are a few issues. Note, we (Ansys staff) can only give guidance so you've got some reading to do. 

      • The mesh is not suitable for CFD. You want at least 10 cells across the channel, ideally more and to resolve the surface in more detail. With the above it'll be hard to tell what's a mesh artifact and what's biological. 
      • The UDF looks OK, but check how it performs. I've looked at it, not checked the code. 
      • Inlet diffusion shouldn't make much difference here, but check the manual for more information. 
      • You may want to review diffusivity coefficients. For liquid systems you're a little off what I'd expect. 
Viewing 2 reply threads
  • You must be logged in to reply to this topic.