In DSMC, in the near continuum regime, an ESBGK collision mode can be used to minimize computing time. Additionally, when using DSMC and the VHS or any other model, the viscosity and Pr can be specified empirically by tuning the reference diameter (molecular diameter cross-section) and so on.... so my question is as follows:
Is there a simple way of specifying the heat capacity instead of resorting to detailed calculations regarding the degree of freedom of a given molecule (translational, vibrational, and rotational energy modes and so on)?.
Calculating the rotational and vibrational energy transfer in collisions is also time consuming (for polyatomic molecules). Plus, I have already tabulated data of Cp vs Temperature, and I would like to use this information at each cell depending on the cell average temperature, ... is this possible? or has this been implemented in an article/chapter/book?
But the purpose of MD simulations isn't that? We don't have to introduce any macroscopic magnitude, they develops autonomously.
Well Yes and no. 1) DSMC is also used when the Knudsen number is above 0.1 (Navier-Stokes-Fourier no longer apply) and 2) "real" molecular diameter aren't used (for engineering calculations), instead molecular cross-section is "adjusted" to yield the correct viscosity at a given temperature.
In my case, I have viscosity and Cp data available VS Temperature, and the molecules have several degree of freedom, which is why I am trying to figure out a "faster" (more efficient) way to implement this. Plus, im working with a gas flow which has solid particles also, that will need a DSMC (hardsphere) scheme anyway... so I thought on going full lagrangian stochastic approach.
Just for the sake of understanding the question, doesn't cp/cv and Temperature appear in DSMC-VHS acceptance algorithm? I've seen it in Bird's book.
They enter the equation indirectly...when one adds the degree of freedom for each molecule. i.e. you apply the partition principle under the assumption that that some "modes" are unlocked as temperature rises (that is, the vibrational modes, bending, etc are either on or off). energy exchanged on collisions are "spread" along the different modes. That way you capture the effect of Cps that are variable with temperature.
Yup, however, my question comes more from an engineering background. Some times you have a correlation for Cp and would like to use it in the DSMC simulation. So my doubt/question goes in that direction.
- Badges
- Science topic