Troubleshooting Expansibility Convergence in Discharge Coefficient Loss Models

Flow rates in excess of the critical flow rate (i.e. supersonic flow rates) or very low flow rates may cause issues with calculating the expansibility factor when using the Cd (Inline) subsonic loss model for orifices or the Cd (Perry's) subsonic loss model for venturis.

Supersonic flow rate guesses are not typically an issue in practice since they result in using the sonic flow model, which does not rely on expansibility factor. There are two exceptions.

  • The first exception is due to a discrepancy between the sonic CdA, specified by the junction parameter CdA (for Sonic Choking - Optional), and the subsonic CdA calculated from the Cd and the orifice or venturi area. This can be quickly resolved by setting the sonic CdA equal to the subsonic CdA.

    • Note: If no sonic CdA is specified by the user when using these loss models, it will be assumed to be equal to the subsonic CdA.

  • The second exception is due to a discrepancy between sonic flow rates as defined by the subsonic flow model (Perry, 10-14 (1997)Perry, R. H., Green, D. W., Perry's Chemical Engineers' Handbook, 7th Edition, McGraw-Hill, New York, NY, 1997.) and the xStream sonic flow model. This may cause issues with steady-state expansibility factor convergence. This discrepancy can be caused by the real gas terms that are present in xStream's sonic choking calculations but not in the sonic choking method from Perry's. If this is the cause, using the ideal gas equation of state may resolve the issue. If it converges with an ideal gas, try using Transfer Results to Initial Guesses and running the model again with the original fluid settings.

For expansibility factor convergence issues related to low flow rates, reduce initial flow guesses in all pipes in the leg.

For orifices, first using the Cd (Exit) loss model and transferring results to initial guesses may aid expansibility factor convergence for the Cd (Inline) loss model. This can also be done with the venturi using a different subsonic loss model such as the K factor loss model with a sonic CdA slightly lower than the known subsonic CdA before changing both CdA values to the known value.

Standard convergence troubleshooting practices may also resolve the issue.

Related Topics

Orifice Loss Model

Venturi

Orifice X is using an exit discharge coefficient loss model but is an inline junction. Consider using the Cd (Inline) loss model

Junction X is not sonically choked, but has a pressure ratio at or below the critical pressure ratio calculated by the Perry's discharge coefficient loss model

Discharge Coefficient Loss Model

Troubleshooting