simtk.org

Hello,

I am attempting to run a rigid wall simulation through SV GUI. I assign a pulsatile parabolic flow profile to my input, assigning the data points from one cardiac cycle of flow as the flow file. Values are negated to represent flow opposite the normal. My RCR values are similar ranges as those in the examples on the SV github. I'm curious what could be happening in my simulations to be causing an output of negative pressures?

Additionally, the velocity output is consistently low. i.e., the flow file maximal value is 22 mL/s, but the simualtion output shows 2.7 mL/s as the maximum velocity value. What could be the issue here?

Best,
Joanne Sarsam

Hello Joanne Sarsam,

Which solver are you using?

Are the nonlinear solves within each time step converging (i.e., residuals are < 1e-3) ?

Is your mesh fine enough to correctly approximate the flow/pressure distribution?

Have you run the simulation for enough cycles that the results have converged (i.e., time-varying results don't changed from cycle-to-cycle).

Cheers,
Dave

Hi David,

I am using the NS solver, and the time steps are converging with proper residuals. I am running a rigid simulation with 5000 time steps, where each timestep is the length of a cardiac cycle (0.435 secs). My mesh is fine, with 249,000 elements. I ran the simulation for 5 cardiac cycles.

I am also getting odd results for my velocity outputs; I prescribed an inflow waveform, parabolic profile, in mL/s, to my inlet boundary condition, based off of Doppler velocity data. However, in the output of my results on paraview, the velocities do not oscillate like in the waveform. Instead, it ranges from 0 to the inflow velocity prescribed for time 0 secs, without ever oscillating. Im not sure if there's some sort of issue in my flow file.

Here is the velocity profile: Here is the pressure profile: Here are my RCR conditions: Attached is my histor.dat file, and my flow file.

Thanks so much for the help as as always.
Best, Joanne

Hi Joanne,

I'm thinking that there must be a problem with the inlet flow. Try running a simulation for a steady flow using a flow file like this and see what the results look like, examine velocity vectors.

Cheers,
Dave

Hello David,

Thanks for the suggestion. I reran the simulation for 3 cardiac cycles, assigning a steady inflow. The velocities I get as output as not close to the flow file values of 100 cc/sec. The velocity vectors also look a bit odd. The pressures are also impractically high, but this is likely due to the fact that we prescribed a non-physiological flow field as input.
Images attached:

Another note -- i notice my pressure waveforms are additive over cycles, rather than stabilizing over multiple cycles. Any idea as to why this might be happening?

Hi Joanne,

It is best to examine velocity using slice planes, easier to see the patterns and such.

The next thing to do is to use the simplest boundary conditions (i.e. resistance) and see if the results make sense.

Cheers,
Dave