[MMatt]

I have simulated particles dissipation inside a room with both the DRWM enabled and disabled, and the results do vary a lot. In my case, it takes twice as much time with DRWM off for the particles to leave the room.

I struggle to see if I should be using DRWM in my case or not. What would be the reason(s) to use it? It is hard to understand if it should be applied to my case or not just reading the documentation.

[vinerm]

If the particle size is comparable to eddy size then particle paths are affected by the eddies, hence, dispersion of particles due to turbulence becomes important and DRW or Cloud Tracking must be used. It will give higher accuracy in that case. If the conditions are not so then you should not use turbulent dispersion.

If the flow is highly nonhomogeneous and diffusion dominated, i.e., low Pe number, then DRW will give unphysical results, however, it does not mean that there is no turbulent dispersion of particles. Its just that DRW is not really valid and cloud tracking should be used.

[MMatt]

Thank you for your answer!

Am I right in saying that if I take the volume-average of K and E over the whole fluid domain, and use Kolmogorov formula for Fluent (https://www.cfd-online.com/Wiki/Turbulence_length_scale) I will get the eddy size (aka turbulent length scale)?

In that case the eddy size is huge compared to the particle size (eddy size 75mm particle size under 1um).

[vinerm]

Volume averaged value for eddy length scale or any field related to turbulence is not a good indicator of turbulence. So, you have to plot the eddy scale contours. Furthermore, if you include minimum three tries for DRW, then the particle tracks with and without DRW should be of similar order. However, if the particles in your domain are of submicron size, then you can certainly neglect the effect of turbulence. Such small particles usually have no drift velocity and go with the mean flow.