Severe pressure oscillations in multiRegionPhaseChangeFlow during bubble growth simulations

[MontelukastNa]

Dear Henning @HenningScheufler , developers @isoAdvector , and foamers,
First of all, thank you for making this excellent TwoPhaseFlow repository publicly available. It has been extremely helpful for phase-change modeling research.

Recently, I have been using multiRegionPhaseChangeFlow to simulate nucleate boiling bubble growth for validation purposes. During this process, I encountered a persistent issue related to pressure oscillations, which seems to depend strongly on the choice of thermophysical properties.

💡 1. Observation with rhoConst + rhoConst case
After reviewing several example cases in the repository, I noticed that many simulations adopt rhoConst + rhoConst thermophysical models.
Using the ScrivenWedge case as a benchmark:
Bubble growth radius (from α field) appears physically reasonable, and the temperature distribution is smooth and consistent
However, the pressure field exhibits severe spatial and temporal oscillations. The pressure field shows strong fluctuations that do not correlate with the smooth evolution of α or T.

📑 Fig 1 p field and T field for the <rhoConst + rhoConst> case
Gif info: 0.1ms/frames 16frames, 15fps

💡 2. Tests with compressible formulations
Since the documentation and related publications indicate that multiRegionPhaseChangeFlow accounts for compressibility, I tested additional thermophysical model combinations:

• perfectGas + perfectFluid
• rhoConst+rhoConst (with eConst)
• perfectGas + rhoConst
• perfectGas + Boussinesq

Case setups (ScrivenWedge.thermoproperties.zip) and animations are attached as follows:

📑 Fig 2. p field and T field for the <perfectGas + perfectFluid> case
Gif info: 0.1ms/frames 16frames, 15fps

📑 Fig 3. p field and T field for the <rhoConst+rhoConst (with eConst)> case
Gif info: 0.1ms/frames 16frames, 15fps

📑 Fig 4. p field and T field for the <perfectGas + rhoConst> case
Gif info: 0.1ms/frames 16frames, 15fps

📑 Fig 5. p field and T field for the <perfectGas + Boussinesq> case
Gif info: 0.1ms/frames 16frames, 15fps

However, the pressure oscillation problem persists in all compressible configurations.

💡 3. Physical consequence: bubble volume oscillation
ScrivenWedge represents bubble growth away from the heating surface. To examine growth on a flat heated surface with a seed bubble, I simulated the 2014 IHMT-Yabuki case with 10.8K. Pronounced pressure instability was observed and can even induce non-physical phase change:

📑 Fig 6. p field and T field for the <perfectGas + perfectFluid> case with 10.8K superheat
Gif info: 0.1ms/frames 57frames, 15fps

In addition, I tested a wide range of boundary condition types (fixedValue, totalPressure, prghTotalPressure, etc.), and the oscillations persist in all cases, suggesting the issue is not strongly related to boundary condition selection.

💡 4. Question
Could you please advise:

• Whether such pressure oscillations are expected behavior in the current implementation?
• Are there recommended thermophysical model combinations for stable phase-change simulations?
• Is there a known strategy to suppress these oscillations (e.g., solver settings, pressure treatment, source-term smoothing, or time-step constraints)?

Any guidance would be greatly appreciated.
Thank you very much for your time and for maintaining this repository.
Best regards
Zhu