Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.
Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
Marangoni Stirring / Phase Change: Weak Contribution Settings
Posted 1 mars 2013, 01:06 UTC−5 Fluid & Heat, Heat Transfer & Phase Change, Modeling Tools & Definitions, Parameters, Variables, & Functions Version 4.3b 3 Replies
Please login with a confirmed email address before reporting spam
I have a problem with a weak contribution in a laser surface melting model:
My model takes into account solid-liquid phase change using an enthalpy-porosity technique and liquid phase volume forces using the Boussinesque Approximation (as in the Continuous Casting tutorial). The model also takes into account Marangoni Forces on the top surface of the material via a weak contribution boundary condition with an expression:
test(u)*gamma*Tx
where u is the velocity x component, gamma is a surface tension/ temperature gradient (a constant) and Tx is the Temperature in the x direction.
The resistive volume force term in the domain beneath the top surface serves to brake the fluid motion down to that of the material (the model is Eularian, meaning the material moves rather than the laser spot) when the material is solid, and allowing additional fluid motion within the molten area.
The trouble is, when the weak contribution is added to the top surface, it seems to allow additional fluid motion in areas where the material is supposed to be solid.
How can I tie the volume force term to the Marangoni weak contribution on the surface to ensure that any additional fluid motion is only allowed within the liquid phase of the material?
Many thanks in advance for your help
Dan
Please login with a confirmed email address before reporting spam
I think your problem is the high velocity gradients at the surface due to Marangoni stresses. Therefore, you will need high braking terms to dampen the velocity in the solid region.
Normally, it helps to adjust your constants in the braking term (as it is used in the tutorial) to give this term a higher priority compared to the inertia of the accelerated fluid. Unfortunately, this usually leads to convergence issues...you can face them by choosing a larger solidification interval for the first calculation, let's say dT=200 K, and then make this interval as small as it should be stepwise by use of the parametric sweep. The other way is to use a fine mesh density but that requires large amounts of RAM, especially for 3D calculations.
Best of luck,
Marcel
Please login with a confirmed email address before reporting spam
Please login with a confirmed email address before reporting spam
Great, thanks Marcel. I have a good feeling that this is the right way forward. Will update the thread to let you know if it works.
Hi, I have the same convergence issue with the same problem, have you figured out how to deal this convergence issue?
Thanks.