[Madz Danzz]

Hello everybody, I am new here and looking for some help from you experts!

For my work I have to simulate closed valve leakage flow (either gaseous or liquid form), on a satellite/shuttle for really small valves (leakage around 10^-6 N cc/s). So it would be the leakage from the upstream duct to the downstream duct trhough the seal of the valve.
I was thinking of starting with some helium (real case is with agressive substances such as NTO/MMH...). The valve has a joint (polymer for example) which is located on the moving part (such as a moppet on a moppet valve) and when closed the joint comes into contact against a seat on the body valve where it deforms. So the leakage is through the "defaults" of the material, due to the rugosity of the surfaces. Does anyone have an idea on how to simulate this? I am a but lost and have not done a lot of CFD

[Martin_Sz]

Simulating closed valve leakage flow, especially for small valves with very low leakage rates, can be a challenging but rewarding CFD project. Here's a breakdown of the key steps to consider for your specific scenario:
1. Geometry Creation:
* Valve Design: Meticulously model the valve geometry, incorporating the upstream and downstream ducts, the movable parts (like the moppet), and the crucial sealing joint. Pay close attention to accurately capturing the surface details, including the rugosity you mentioned, which significantly affects leakage.
2. Material Properties:
* Solid Materials: Define the material properties for the valve body, moppet, and joint. You'll likely need elastic or elastomeric material properties for the joint to account for its deformation under closing pressure.
* Fluids: Specify the fluid properties (helium for initial simulations, then NTO/MMH for actual conditions) considering temperature variations if applicable.
3. Boundary Conditions:
* Inlet/Outlet: Set pressure or mass flow inlet conditions at the upstream duct, and a pressure outlet condition at the downstream duct, reflecting the pressure difference across the closed valve.
* Walls: Define no-slip wall conditions for all solid surfaces.
4. Mesh Generation:
* Crucial for Leakage: A very fine mesh is essential, especially around the sealing joint, to accurately capture the small leakage flow paths.
5. Solver Settings:
* Viscous Flow: Since you're dealing with low leakage rates, a viscous flow solver with appropriate turbulence modeling (e.g., laminar or RANS with low-Reynolds number corrections) might be suitable for your initial simulations.
6. Leakage Rate Calculation:
* Monitor Mass Flow: While simulating, monitor the mass flow rate across a designated surface representing the leakage path between the ducts. This will provide the quantitative leakage rate value.
Additional Considerations:
* Validation: If possible, compare your simulation results with experimental data (if available) to validate your model.
* Multiphase Flow (NTO/MMH): For NTO/MMH simulations, you might need to consider multiphase flow modeling with appropriate property models for these propellants.
Remember, CFD simulations involve iterative refinement. Start with a simplified helium case to gain confidence in your model, then gradually increase complexity towards the actual NTO/MMH scenario.

[Madz Danzz]

Thank you for your answer, really appreciate it!
Could I ask you why do you think it is necessary to represent the whole valve and not just the part where the sealing takes place to simplify the model? I thought it would be simpler this way.
And furthermore, depending on the type of rugosity, as you said, the leakage will be very different. Do you know how to actually represent the rugosity on solidworks or on ansys for example (or other program)?
For example I have some leakage data for a valve to which I could compare but I have to represent its rugosity which is of around 1 micron ( a bit less) and circular pattern (this way more difficult for the fluid to exit through the channels than if it was radial/random pattern). I am not an expert on either of these fields so feel free to correct if I am saying nonsense.