Fluids

Fluids

Incorporating the compressibility of blood using the SST k-omega model

    • Vincent
      Subscriber

      Hello all,

      I am currently working on a CFD simulation of blood flow and I am trying to incorporate the physiological property of blood being compressible. From my research, I have found that the Non-Newtonian Power Law model is commonly used to represent the behavior of blood as a non-Newtonian, compressible fluid. However, I have also found that this model is not available for turbulent flows in my simulation software.

      I am currently using the SST k-omega turbulence model in my simulation and I am wondering if there is a way to incorporate the compressibility of blood while using this turbulence model.

      I have searched the internet and forums for information on this topic but have not found a clear answer. I would greatly appreciate any guidance or suggestions on how to incorporate the compressibility of blood in my simulation while using the SST k-omega turbulence model. 

      Thank you in advance for your help.

    • DrAmine
      Ansys Employee

      I think you want to include the Non-Newtonian property of blood with turbulence model. In turbulent flows the laminar viscosity will be several magnitudes smaller than the the turbulent / flow viscosity. In transition regimes both viscosities might be of similar magnitude and one require a good tranisiton model.

      Is your case characterized by rather high blood "laminar" viscosity?

      Anyway you can enable turbulence experts settings and be able to sue turbulence modeling on top of a non-newtoninan viscosity model by using turbulence expert option in TUI under define models viscous.

    • Rob
      Ansys Employee

      I didn't think blood was compressible?  Typically, blood flow is laminar, if it's turbulent (ignore the medics, they don't know what turbulent means from an engineering stand point) you'll damage the red cells and that's (the medics understand that bit) bad. You then use the best nonNewtonian model for viscosity. 

    • Vincent
      Subscriber

      Hi, thank you both for your reply. Indeed, I would like to incorporate the non-Newtonian property of blood while using a turbulence model. The non-Newtonian property shows that viscosity should decrease with higher shear rate, so a lower viscosity with higher presence of turbulence? I was thinking about incorporating compressibity, as the red blood cells introduce this property of blood, slightly, but it's neglectable in most cases. I will ignore it in my case. 

      I'm simulating the filling phase of the left ventricle in which turbulence is present. I tried using a laminar model but the residuals did not converge (steady state simulation). Also from literature I see that the SST k-omega model seems to be applicable for this case.

      So I want to introduce the non-newtonian blood property for an SST k-omega model, maybe with the non-newtonian power law model? But this is disable for the SST k-omega model. Why is this normally disabled? (is it incorrect to model fluid as non-newtonian for the turbulence model?)

       

    • DrAmine
      Ansys Employee

      It is disabled as it R&D. You can enable it as mentioned in my post. In a real turbulent case the laminar viscosity will be much lower than the eddy viscosity.

    • Rob
      Ansys Employee

      Erm, it's not turbulent, at least not in the engineering sense. Medics (and geologists) have a slightly different definition for turbulent flow. They assume (incorrectly) that eddies in the flow imply turbulence, rather than being laminar eddies. What you're seeing is (probably) eddy formation from part of the geometry, and you may want to read up on DVT and aneurysms for why that may be bad. Convergence may be a little poor as the flow may be inherently transient - please post a screen shot of the residuals. 

      If you really want turbulence you can use a UDF as that bypasses the checks: they were added in v6.0 because I tested it and didn't know you couldn't have turbulence & nonNewtonian flow at the same time, at least not reliably with the models we had. There is low Re option, under 

      /define/models/viscous/turbulence-expert  turb-non-newtonian?

      but you MUST read the manual pages (if they still exist) on why it's often not a good idea. 

Viewing 5 reply threads
  • You must be logged in to reply to this topic.