[gzotin]

Hi everyone,

I'm trying to model textile ducts in a large room, and I'm having some problems at the convergence level. I've already done two simpler simulations: only the pressure drop over the fabric and the other for a part of the cylindrical duct in a very small room. They work fine, and I could get to the convergence, but when I try the larger room with longer ducts (and their connections between themselves), I can't get the convergence, and even from the first steps I have already high residuals (usually from k and eps).

For the simpler simulations, I used porous media and porous jump, for comparison. But for the larger one, I decided to use porous jump (even if I know that this method is harder to converge). The thickness of the fabric is 2 mm.

I've already tried to discretize the contacts between the fluid regions (with some inflation layers) and I'm patching the regions for a better initialization.

I'm looking for a piece of advice

BCs

• 2x mass flow inlets (one in each side of the room)
• 2x mass flow outlets
• 2x pressure outlets
• 2x porous jump (for each duct)
• general walls

Initialization
Standard initialization computed from one of the inlets and patched regions.
Patched regions :

• Pressure inside the ducts: 98 Pa
• Pressure inside the room: 0
• Velocity inside the ducts from the inlet bc