[derkaiser]

hello,

there has been en experiment of a fluidized bed with 36500 glass beads of diameter 2,5mm which we are trying to simulate with the eulerian multiphase model.
there has been an euler-lagrangian simulation which fits the experiment quite well. you can see a comparison between experiment and euler-lagrange simulation here:
http://www.est.tu-darmstadt.de/index...am-est/dem-est

in the eulerian model we define the glass beads as a solid fraction with density 2556 kg/m³.
problems arise in modelling the viscosity of the solids phase. using either schaeffer or lun-et-al model for the frictional viscosity leads to fluent limiting the turbulent viscosity ratio to a value of 1*10^5. i assume that this leads to a underprediction of the viscosity of the solids fraction.
a plot of the viscosity shows that the viscosity in most areas of the flow field is zero. for the given mass flow of air from the experiment the particles become thrown out too far from the bed. this effect can be reduced by reducing air mass flow which is not desired.
the problem is now how to model the viscosity in a right way or unlimit the ratio of the turbulent viscosity ratio? values that are of interest are the frictional viscosity, angle of internal friction etc.

thanks in advance for your help!

[sangramroy]

hi,

Kindly check your grid once more. Too much skewness can lead to the problem as you mentioned. Other wise you can chnage the under relaxation factors for turbulent properties.