Notice: The Ansys Innovation Space websites will be temporarily unavailable on Monday November 6 from 8:30 EDT to 1:00pm EDT for scheduled maintenance.
We apologize for any inconvenience this may cause and appreciate your understanding.
Multiphase Flows

Multiphase Flows

Solver message during DPM calculation: “number of stepsize underflows during particle integration step is x”. What does it mean and how to get rid of it?

    • FAQFAQ
      Participant

      These messages are thrown when the ODE is being solved under the coupled heat mass option. It is a counter to monitor how many times, the integration error stayed above the default tolerance. They typically show up, when the Runge-Kutta solver we are using is taking too much steps to integrate the underlying ODEs for a given particle integration time step. Since this message has been issued in the past for each integration step which fails we are collecting them now and issuing only once per DPM Iteration. One workaround is to use a different ODE solver. If you are running multi-component droplets then there is an rpvar which must be set to #f. (rpsetvar ‘dpm/rkmulticomponent? #f) For pure droplets you must set (rpsetvar ‘dpm/rkvaporization? #f) After that you must issue the command (dpm-parameters-changed) These RPvars switch to an Eulerian integration of the ODEs. Another workaround is trying to understand what is causing these messages. Typically this also indicates wrong settings, either of the DPM properties or of the numerical settings. Therefore, user should check thermodynamic consistency of DPM properties like latent heat of vaporization, specific heat capacities of gas and liquid, vapor pressure, etc. Another workaround might be to reduce the particle integration time step by disabling accuracy control and automated tracking scheme selection. In some cases this might lead to large integration time steps which might be efficient for the hydrodynamics but not for integrating the heat and mass transfer equations. By the way, since v14.0 we do not need to use these rpvars anymore. They are behind the Coupled Heat Mass solution in the numerics tab of the DPM panel model.