Altair RISE

A program to recognize and reward our most engaged community members

Nominate Yourself Now!

Zero gradient boundary in Acusolve

Hi experts,

I am simulating a case involving a variable-density fluid propagating in a water tank, and I need to set an outlet at the end of the tank. I tried using a pressure outlet boundary condition (considering hydrostatic pressure), but the results were not very good. Now I want to use a zero-gradient boundary condition instead. I’d like to know how to set this up in Acusolve. Is it done through the element boundary condition settings? If so, how exactly can I configure it? I couldn’t find the relevant instructions in the manual.

Thanks a lot

Find more posts tagged with

AcuSolve

Accepted answers

All comments

acupro

I'm pretty sure the standard outflow BC results in a zero-gradient condition across the boundary.

Can you give more information about your simulation - including images? And maybe include/attach the input and Log files?

Ken7713

Thank you for your response. I've attached images of the case, the .inp file, and the .log file. I noticed that in the .inp file, the outflow section has a parameter called non_reflecting_factor with a default value of 0. Should this be changed to 1? Additionally, since I'm using the Simlab GUI, I can't find where to set this parameter.

Files.zip

acupro

This is an interesting case. You have an outlet - but no inlet. What are you assuming about what is outside the simulation domain?

Do you assume the tank is much longer - but you want to cut the length of the simulation domain? You might be better off deciding what is your domain length of interest for results - then make the simulation domain quite a bit longer to push the outlet effects away from your region of interest?

Do you assume the tank ends? Then really all the water (of both 'types') would fall out of the tank and the tank would fill with air. This would then be multiphase to capture the air rather than with species - but would not be possible with AcuSolve as it would be three fields - water, saltwater, air.

It's interesting as the mass-flow out and in become the same value - that value increases to roughly 8 kg/sec (time step 3862) before decreasing again (about 3.0 kg/s by step 4370 where the Log file ends). This just might not be a stable problem with these assumptions.

You can try increasing the non-reflecting-factor, but I think that's more for pressure - to prevent pressure waves bouncing back and forth. You would need to manually edit the input file, then run from command-line/command-prompt or Compute Console using that edited input file.

You could also try changing the species backflow type - maybe try area_average or exiting_area_average to see if that helps stabilize things.

species_1_back_flow_type or species_1_back (enumerated) [=value]

How the constant species_1 boundary condition is defined under back flow conditions. Used with outflow and far_field types with back_flow_conditions=on.

value Specified value. Requires species_1.
area_average or area_ave Average of solution over entire outflow surface. exiting_area_average or exit_area_ave Area average of solution over subset of outflow surface that has flow exiting the domain.
mass_flux_average Mass flux average of solution over entire outflow surface. exiting_mass_flux_average Mass flux average of solution over subset of outflow surface that has flow exiting the domain.

You could also try increasing the mesh density - to see how much more mesh affects the solution(s) of interest.