[avd28]

Hi all,

I'm trying to simulate a biological fluid flow which requires the use of pulsatile sinusoidal wall motion of the mesh. I have the UDF compiled and running. However there is a problem.

Each node is supposed to deform with the equation of the form Asin(wt) where t is the current time.

However, when the time step size is decreased from say 0.1 secs to 0.01 secs, the mesh amplitude unaccountably increases, for the same number of iterations. This of course results in formation of negative volumes in the mesh and the simulation crashes.

I have attached the udf file and will provide a dropbox link to the case file. In order to view the problem, simply simulate with a time step size of 0.01 as compared to 0.1 secs.

https://www.dropbox.com/s/pjdkjink6t...test1.cas?dl=0

[`e`]

The case file and UDF you have posted are not changing the mesh at all (I've tried with ANSYS 15.0). Is the 'Wall' boundary supposed to encompass both the cylinder wall and spherical shape? Lastly, because you're dealing with small numbers (micrometer), I recommend using double precision (you're currently using single precision).

[avd28]

Hi,

I'm not sure, why that is the case. I'm using ANSYS 15.0 as well. Have a go with this one.

https://www.dropbox.com/s/1bui77s9ea...based.cas?dl=0

Yes, your assessment about the wall is correct, the entire assembly should deform sinusoidally.

I'll incorporate your point about double precision.

Thanks

[azi]

Hi, Have you solved the problem?
I am facing a similar problem. I need to insert a displacement profile which is a function of the time to the wall of the artery. It couldn't figure it out how to incorporate it.
Thanks for your help.