code_saturne User's Forum

Hello,

in USLAG1() the parameter INJCON enables the continuous injection of particles during a lagrangian time step.

This is done by shifting the position of new particles in the grid by  alpha*timestep*velocity, where alpha is a random number in the range 0 ..1.

This approach is fine. But I doubt it really works, because I can't find the code line where the velocities are initialized correctly.

The sequence of subroutine calls and actions in LAGENT():

• - USLAG2() -> setting of boundary conditions
  - call either LAGNEW() or LAGNPR()  -> determination of the face/cell where the particles are created
  - INJCON .eq. 1 then LAGNWC() -> shift positions of particles
  - initialize particle velocity  (array ETTP) with the values from the array RUSLAG
  - USLAPR() -> setting of velocity/diameter/weight profiles by the user

To clarify, in lagent()/lagnwc() the particle velocity is used for the calculation of a position shift before it is initialized.

As far as I understand the code the call to lagnwc() must be placed after the velocity initialization and after the call to uslapr() with idvar = 1.

Kind regards,

Ralf Becker

Hello,
You're right. It's  a mistake. We will correct it.
Best regards
Mickaël Hassanaly

Previously Mickael Hassanaly wrote:

hai Mickael
i am using code-saturne for particle tracking....
i have a trouble in solving in it................. that is when the particle count has been increase from 1000 to 100000 the time taken to remove the particles in both case is same.
so please reply if you know the solution,
Then how to get the particle count at different time steps,..... Then how to have different size of particles in one injection
with regards
thanking you
thejaraju.r

hi mickael
i am working on particle tracking............
i need to know how to have a non-uniform distribution of the particles in inlet.........
that is particle of different diameter at same injection point
thanking you
with regards
thejaraju.r