Home >> Forums >> Particles settling velocity

Particles settling velocity

 
Hello ! 
I would like to know how the particles settling velocity is calculated by ICHTHYOP in the 3.3 version?
I am trying observe some particles spreading in a 3D space but my particles seem to reach the bottom in a few hours, usually 2-3 hours, (which is pretty short for a 80 days long model). My model is runing with the buyoancy and growth parmaters activated but settled so that both will stay constant. 
 
I tried to run a few tests while varying my density value and my particles reach some amazing settling velocities ! The velocities have been calculated as instaneous velocities using ICHTHYOP output using a simple relation u = (z2 - z1) / dt. and reach some velocity values like 0.35m/s ...
 
I am aware that it can as well be an problem regarding the ROMS input but I already checked it and it seemed good. It is also a possibility that there is no problem with ICHTHYOP output but I would like to make sure before going on with the rest of my projects.
 
thank you ! 
Grégoire
 

Ichthyop version: 
Ichthyop 3.3alpha
Hydrodynamic dataset: 
ROMS
Java version: 
Java SE 8
Operating system: 
Mac OS X

Hi Gregoire

I have never personally used the Buoyancy module.

But here is what I was able to find from the comments in the code:

     * dw = vertical velocity [cm/second] due to buoyancy
     * g = acceleration of gravity [cm.s-2]
     * d = minor axis of the ovoid [cm]
     * l = major axis od the ovoid [cm]
     * rhoParticle = egg density [g.cm-3]
     * rhoW = sea water density
     * deltaRho = rhoParticle - rhoW
     * mu = water molecular viscosity [g.cm-1.s-1]
     * dw = 1/24 * g * d * d * deltaRho/rhoW * (1/mu) * ln(2 * l / d + ½)

with
   
    final private static double MEAN_MINOR_AXIS = 0.05f;
    final private static double MEAN_MAJOR_AXIS = 0.14f;
    final private static double LOGN = Math.log(2.f * MEAN_MAJOR_AXIS / MEAN_MINOR_AXIS);
    final private static double MOLECULAR_VISCOSITY = 0.01f; // [g/cm/s]
    final private static double g = 980.0f; // [cm/s2]

The output of this function is then converted into meters to fit Ichthyop depth units.  Temperature should be in Celcius and Salinity should be in PSU.

If you want me to have a look at why you have such values, I invite you to send me your configuration file and the NetCDF files you use (a few time step is enough).

Nicolas

Hello Nicolas, 
 
I am thankful for your answer, the formula you gave me seems to be the same as the one given by Parada and al. 2003.
You will find my configuration and the NetCDF file I used in the attachments. I cut the NetCDF file to 4 timesteps, will that be enough ? 
Concerning the configuration, I am sending you the one I used to run some tests so the only discrepancies are on the time scale and the sampling frequency as I wanted to have look on the sinking velocity.
 
Grégoire
 
 

Hi Gregoire

It seems that your NetCDF file is corrupted (I have only 0s for all variables).

Nicolas

 
I am sorry about this ... I am still not used to manipulate NetCDF files ... Here is the file with less timesteps than before so that the size is not too big for the forum. I checked the file and the values seem to be in without problem.

Hello Nicolas, 
 
I think I have found were my errors come from !
According to Parada et al. (2003) - the research paper from which this formula comes from according to Lett et al (2008) A Lagrangian tool for modelling ichthyoplankton dynamics - and the part of the code you send me there are two problems with my configuration. 
 
Firstly I can't change the dimension of my particles : they are considered as prolate spheroids following the dimensions you gave me previously (major axis = 0.14cm and minor axis = 0.05cm) which is way bigger than the particles I am studying ...
 
Secondly this formula can only be used for Reynolds numbers inferior to 0.5 which mean your particles density has to be included between 1.021 g.cm-3 and 1.038 g.cm-3
 
It is possible that I missed some other parameters but my configuration should have been enough to get completely out of range from the formula used in the source code. I will have to find another way to pursue my research ! Thank you for your help. 
 
Grégoire