Ill-Conditioned LHS Matrix with Small/Zero Inflow

[fbelik]

I have experienced ill-conditioned LHS matrix errors/warnings when running the example Cylinder Project provided on the SimVascular webpage when I change the inflow file. The provided steady inflow.flow file runs fine without issue. (Running the December 2025 version on Ubuntu 24 and Ubuntu 25)

Here are four inflow files.

1. cos_flow_shifted.txt: Oscillatory flow shifted away from zero, runs fine.
2. zero_flow.txt: Zero inflow, results in a zero matrix error.
3. sin_flow.txt: Oscillatory inflow file that starts with positive inflow and eventually goes to zero, of the form $-50 + 50 \sin(2\pi t)$. Results in an ill-conditioned LHS matrix error.
4. cos_flow.txt: : Oscillatory inflow file that starts with zero inflow of the form $-50 + 50 \cos(2\pi t)$. Results in an ill-conditioned LHS matrix error.

I have been communicating with @ktbolt about this and how the new version will report a warning instead of an error. I am waiting for the new Ubuntu release to test this further. I believe I experience the same issue with an RCR outlet condition.

Is there a reason these test cases would result in ill-conditioned matrices? The provided inflow file, inflow.flow, which is steady 100 inflow, does not have ill-conditioning occur.