[Siba11]

Hi all.

I need to model blood clots inside stenosed arteries.

I found a similar paper that simulated blood clotting using lattice-boltzmann method. The authors injected a 'passive scalar' at every lattice node during each time step. Convection-diffusion eqn was solved for this massless scalar. If, at any lattice point, the concentration of this scalar exceeded a threshold, then that lattice point was changed from a "fluid node" to an "obstacle node" & solid surface Boundary cond. was applied. Thereby mimicking a solid clot.

I'm curious if it is possible to achieve something like this in ANSYS FLUENT (possibly via a UDF).

Thank you.

Link to original paper: https://ieeexplore.ieee.org/abstract/document/1524320

[Siba11]

Hi all.

I tried out something. Although it didn't work correctly but i'm posting it. Maybe it'll help someone come up with another idea.

Step 1: I switched on the UDS transport eqn.
Step 2: Then I entered a value for UDS diffusivity & source term (in the cell zone).

This will help me monitor the concentration of the passive scalar.

Step 3: I compiled a small DEFINE_ADJUST udf to set the cell velocities=0 in those cells where the UDS concentration exceeded a threshold.

It worked for a few iterations but then the residuals blew up.

[p.norouzi.gv]

Were you able to model it in Fluent?

[lolno]

Originally Posted by Siba11

Hi all.

I tried out something. Although it didn't work correctly but i'm posting it. Maybe it'll help someone come up with another idea.

Step 1: I switched on the UDS transport eqn.
Step 2: Then I entered a value for UDS diffusivity & source term (in the cell zone).

This will help me monitor the concentration of the passive scalar.

Step 3: I compiled a small DEFINE_ADJUST udf to set the cell velocities=0 in those cells where the UDS concentration exceeded a threshold.

It worked for a few iterations but then the residuals blew up.

Nice sharing Siba11, just wanted to ask does this method will work with porous media?

[Siba11]

Originally Posted by p.norouzi.gv

Were you able to model it in Fluent?

Hi, yeah I was able to model it in Fluent. I used cell registers to mark the mesh cells which exceeded the threshold. Then, I separated and deactivated those cells.