Hello,
I am simulating a CVD (Chemical Vapor Deposition) reactor and would like to know if Nek5000 is suitable for this type of problem. Below are the details of my case:
Problem description:
I have hydrogen gas (H₂) flowing through a heated reactor at sub-atmospheric pressure. The gas enters cold and gets intensely heated by the reactor walls and a rotating wafer surface. The key challenge is that pressure must be treated as a thermodynamic variable — density changes significantly due to heating, not due to high-speed flow.
Operating conditions:
Operating pressure: 200 mbar (20,000 Pa)
Inlet temperature: 300 K
Wall temperature: ~1600 K (heated by induction coils)
Inlet velocity: 2 m/s
Working gas: Hydrogen (H₂)
Mach number: ~0.0015 (deeply subsonic)
Reynolds number: ~36 (laminar)
Density variation: ~5× (0.016 kg/m³ at 300 K to 0.003 kg/m³ at 1600 K)
Rotating wall boundary (wafer at 6.62 rad/s)
I believe Nek5000's low-Mach formulation would be naturally suited for this regime since it removes the acoustic CFL constraint and avoids the pressure-density coupling issues I experienced in OpenFOAM. However, I am unsure about the feasibility of implementing multi-species reacting flow.
Geometry:
The reactor has a converging-diverging channel feeding into a rectangular reactor section (~0.15 m × 0.03 m × 0.287 m) with heated walls on all sides and a rotating wafer on the bottom wall. The mesh is currently ~8 million hexahedral cells in OpenFOAM.
Any guidance, references, or example cases would be greatly appreciated.
Thank you.
Hello,
I am simulating a CVD (Chemical Vapor Deposition) reactor and would like to know if Nek5000 is suitable for this type of problem. Below are the details of my case:
Problem description:
I have hydrogen gas (H₂) flowing through a heated reactor at sub-atmospheric pressure. The gas enters cold and gets intensely heated by the reactor walls and a rotating wafer surface. The key challenge is that pressure must be treated as a thermodynamic variable — density changes significantly due to heating, not due to high-speed flow.
Operating conditions:
Operating pressure: 200 mbar (20,000 Pa)
Inlet temperature: 300 K
Wall temperature: ~1600 K (heated by induction coils)
Inlet velocity: 2 m/s
Working gas: Hydrogen (H₂)
Mach number: ~0.0015 (deeply subsonic)
Reynolds number: ~36 (laminar)
Density variation: ~5× (0.016 kg/m³ at 300 K to 0.003 kg/m³ at 1600 K)
Rotating wall boundary (wafer at 6.62 rad/s)
I believe Nek5000's low-Mach formulation would be naturally suited for this regime since it removes the acoustic CFL constraint and avoids the pressure-density coupling issues I experienced in OpenFOAM. However, I am unsure about the feasibility of implementing multi-species reacting flow.
Geometry:
The reactor has a converging-diverging channel feeding into a rectangular reactor section (~0.15 m × 0.03 m × 0.287 m) with heated walls on all sides and a rotating wafer on the bottom wall. The mesh is currently ~8 million hexahedral cells in OpenFOAM.
Any guidance, references, or example cases would be greatly appreciated.
Thank you.