Fluids

Fluids

High pressure gas delivery into a different medium

    • Carlos Manglano
      Subscriber

      Hi all,

      I'm trying to run a simulation for gas pressure depletion from high-pressure source into an 1atm outlet medium, though small channels.

      High-pressure source: hidrogen/oxygen/helium gas at >>75bar.

      Small channels: hydraulic diameter <5mm, lengths <3cm

      Outlet medium: 1atm Air

      Simulation not starting due to bad definition of BCs, or suddenly diverging before 10 iterations.

      I have tried steady and unsteady, only Air, starting with small delta-pressure between inlet and outlet, velocity-inlet, mass-flow-inlet, pressure-inlet. Always pressure-outlet. Cannot get a converged solution. If it's physical or not, I will analyzed when I get something to work with :) .

      Could someone give some hints on BCs, solution and models/methods best fitted for this simulation, please?

    • Rob
      Ansys Employee

      If you're emptying a bottle (for example) you'd need to patch the pressure in that zone and run transient. For a "large" vessel you may be able to assume a constant inlet pressure somewhere. 

      Typically, you will need suffcient space for the flow to establish, and any jets/plumes to form. You may also need to be very careful with initial conditions given the relatinship between pressure, temperature and density. 

    • Carlos Manglano
      Subscriber

      Thank you Rob,

      The phenomena we want to detect and where it happens, occurs very soon in the depletion process and we can assume constant inlet pressure, thank you for pointing that out.

      The problem with establishing the flow, is that we need the simulation volume to be as accurate to reality as possible. But, I could start with a long inlet pipe to obtain a converged flow locally, cut the inlet domain and map properties at inlet BC.

      Could you be more specific with initial conditions and relationship between P, T and rho, please? 

       

    • Rob
      Ansys Employee

      I can't be too specific, but at 75 bar I hope you're looking at an ideal gas density! 

    • Carlos Manglano
      Subscriber

      That would be another story :) ... but we need to keep things as simple as possible for now.

      So, you cannot throw some hints on how to treat the time steps, the relaxation factors, and those kind of numerical factors, to get a stable start of simulation?

       

    • Rob
      Ansys Employee

      If it's fixed density use a velocity inlet, patch the jet in (carefully) and it may work. However, I suspect 75 bar will choke so you may need to rethink the inlet condition. 

      Time steps should be set such that the flow crosses a cell in a few time steps. Otherwise most of the default solver settings will more-or-less be OK. Have a look in the tutorials, examples and Learning (top left here) for more tips.

      We (Ansys staff) are not allowed to give specific engineering advice in an open forum: others don't have that restriction. 

    • Carlos Manglano
      Subscriber

      Thank you Rob,

      last question ... if I start with a steady solution, how do I model the few-time-steps per cell? Starting with unsteady is a bit strong.

      I know I have choking ... that's my everyday pain.

      cheers!

       

    • Rob
      Ansys Employee

      If it's steady you don't need to worry about time step. Time scale factor may want to drop, try 20-40 and see how you go. You will need to carefully patch in the jet to get the model going. 

Viewing 7 reply threads
  • You must be logged in to reply to this topic.