BC order overwrite?

[wfullmer]

I’m updating our Patil test case for the PIC VVA doc. At the bottom of the bed is a distributor plate modeled with a typical mi. But there is a center pipe that is not in use and should be identified as a dead zone with no mi. In the old inputs I have from years ago, we had two bc.regions: a mi covering the whole face (labeled 1 below) and a smaller region (labeled 2 below) which was also a mi but had zero velocity.

Is placing bc 2 later in the inputs file sufficient for it to overwrite the subregion of bc 1? Or do I need to do something like specify four bc.regions as shown below?

[RobertoPorcu]

Have you tried running the test with the two different BCs configurations? Are the results in agreement or do you see differences?

[wfullmer]

I’m getting (close to) what I want with the top metho above of overwriting bc 1 with bc 2. Jordan mentioned at lunch that it should overwrite based on the order placed in bc.regions not where the actual bc. information is placed in the inputs file. I figured out that the reason I wasn’t seeing a nonuniform profile like this earlier is because I didn’t extend my box through the bottom of the domain so I was applying a bc on a covered 1e-15 sliver outside the valid region.

[wfullmer]