Learning Forum

[valentin.melot]

Hey all!

I have a very simple setup where a single constant temperature water vapor bubble is immobile in a constant temperature quiescent fluid. This is done to study a custom heat exchange law between an immobile bubble and the fluid I implemented using the DEFINE_DPM_HEAT_MASS UDF. The simulation is transient, as is the particle tracking. When turbulent dispersion is disabled, the bubble is immobile, Fluent's reported particle Re number is 0, and tp->Re = 0 inside the UDF (tp is a pointer to the Tracked_Particle structure of my bubble). However, when turbulent dispersion is enabled (with random eddy lifetime, and the number of tries, which is locked, equal to 1) everything stays the same, except that tp->Re is NOT 0 anymore. But Fluent's reported particle Re number is still 0, and the particle is still immobile, with 0 velocity. Why would there only be a difference in tp->Re?

Thanks a lot.

[valentin.melot]

Nevermind, I think I understand why. I think that the fluctuating velocity component added to the particle's velocity is seen when fetching the particle Re number "instantaneously" using tp->Re, whereas Fluent's reports, since I use RANS, contain the mean velocity, which is 0.

[Rob]

Possibly, if you set the stochastic tracking to 10 (as an example) how do the 10 values compare? 

[valentin.melot]

Thanks for answering. I can't put it to 10 since I use unsteady particle tracking. Anyway, I have a question regarding the Discrete Random Walk model. In the screenshot below, Fluent defines the integral time scale T, and it depends on the fluctuating component of the particle velocity (u'_p). How is the latter computed? Since u_p is found by solving the ODE of dx_p/dt = u_p and the force balance on the particle, how is there a fluctuating component? I'm trying to describe clearly the model, and especially why the number of tries is fixed to 1 for unsteady particle tracking.