Why does pressure error do not decrease as mesh is refined? (Stokes flow)

Hello,
Thank you so much for taking the time to respond to my post.

  1. The facets are as expected.
  2. The pressure solution has its largest error near the bottom of the domain, almost like the outflow condition for u makes the pressure struggle. This is what the error looks like at the most refined level:
  3. I personally struggle (a lot) with changing between coordinate systems. Since I have to run various more complicated examples in cylindrical coordinates and compare to Cartesian, I wanted to prevent future mistakes by always writing 3x3 tensors, as the cylindrical tensor has an extra term. I could also change the variational formulation to include that term, so it’s really the same. I should have made them 2D for the MWE, sorry about that.

I finally ran an incompressible problem, and you’re right. The error for the velocity and the pressure now behaves as needed. I based the manufactured solution on this post in steady state. I was aware that there are some issues for compressible flow, I was focused on other aspects of my solution and didn’t connect the dots.

Thank you!