Flow simulation of glassfibre-filled polymer melt

Dear CFDEMers,

I would like to simulate the flow of a viscous (or even visco-elastic) fluid filled with glassfibres (about 1-4 mm in length, no spheres). Is that possible using OpenFOAM and LIGGGHTS? If not, are there any plans to include this feature?

Thank you in advance!
Thomas

Forums:

CFDEM®coupling - User Forum

cgoniva | Tue, 11/29/2011 - 11:58

Hi Thomas,to date there is

Hi Thomas,

to date there is no "visco-elastic-fibre-CFD-DEM_solver" ;-)
,but that is surely feasible within the cfdem coupling!!!

I would suggest to approach the problem step-wise:
e.g. viscioelastic solver with spheres , then go for non spherical particles, then go for fibres...

for viscoelasticity, please see publications and work by Favero (in -ext version):
Viscoelastic fluid analysis in internal and in free surface flows using the software OpenFOAM
J.L. Favero, A.R. Secchi, N.S.M. Cardozo,., H. Jasak

Concerning the plans: for viscoelastic solver there is currently no planned development, non-sphericity is definitely a topic to be covered in future (next year). I doubt that a simple non-spherical this model will be able to capture very long fibres...

Hope that helps you further and gives some motivation.

Cheers,
Chris

KoppTh | Wed, 11/30/2011 - 13:12

Hi Chris, thank you for your

Hi Chris,

thank you for your answer. I agree with you suggesting a step-wise approach. For me visco-elasticity is not the most important topic, but I would like to couple CFD and DEM for modelling the flow of a very viscous polymer melt filled with fibres.
As a first step spherical particles are okay for me. If a non-spherical model was developed, how long do you think the fibres could be? I am working with different kinds of fibres which can be classified as short (max. 1mm) and long (max. 4mm).
At the moment I am doing some rheological measurements in order to obtain good material data for simulations in the future. Also, I will try to determine the distribution of fibre lengths, because the material is not filled with fibres of equal length. Is it possible to define size distributions in LIGGGHTS?

Best regards,
Thomas

cgoniva | Wed, 11/30/2011 - 17:55

Hi! >> Is it possible to

Hi!

>> Is it possible to define size distributions in LIGGGHTS?
yes, no problem.

>> how long do you think the fibres could be?
length is not a problem, actually length to with ratio will cause problems. As this will badly be pictured by a sphere, and you will need to go for e.g. spheres glued together.