WARNING: 0.5241% of the total discrete phase mass was not tracked for the expected residence time

  • Last Post 03 November 2018
  • Topic Is Solved
MohAgha1 posted this 02 November 2018

Dear All;

I was running a DDPM-DEM simulation of the fluidized bed and I got the following warning which causes non-realistic results and divergence in the solver, here attached the warning and screen-shots

hope somebody can give some good recommendation

Order By: Standard | Newest | Votes
abenhadj posted this 02 November 2018

To get away with the warning: Adjust your tracking parameters (Increase Max Number of Steps for example). Moreover as you are trying to calculate the collisions your particle time step size has to be at leat 1/10 of the collision time. You need then to approximate the collision time based on the spring constant, reduced mass and coefficient of restitution.

Best regards, Amine

  • Liked by
  • MohAgha1
abenhadj posted this 03 November 2018

It does not make a sense increasing the number of steps. Check the collision time as mentioned. Regarding mass flow from MATLAB I won't comment. You have the possibility of volume injection in Fluent. Use it meanwhile switch off flow. After the particles settled down switch on flow. You do not require other non drag forces for solids. Use DDPM for dense beds.

Best regards, Amine

  • Liked by
  • MohAgha1
MohAgha1 posted this 02 November 2018

Dear Mr. Amine;

Thanks for your response.

I tried reducing the dem time step to 1e-5, the fluid time step to 1e-4 and increased the maximum number of DPM time steps to 50,000.

Unfortunately, the warning still exists and there is a non-realistic solids void fraction of 0.002 ?!

Can the problem in the parcel's mass flow?

I calculated it into Matlab m-file as follows:

total_mass_flow   = 0.3*0.3*0.01*0.6*density; % mass of the solids occupied 0.3x0.3x0.01 2D domain with 0.6 initial void fraction

m_p         = (pi/6)*diameter*diameter*diameter*density;  % physical particle mass

n_p         = total_mass_flow/m_p/x_count/y_count/z_count;  % total number of physical particles/parcel

mass_flow = n_p*m_p/1e-8;         % parcel mass flowrate in 1e-8 injection time period