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.

welding simulation

Kartikeyan Ramachandran
hi,
i am trying to process model tig welding of metal. i solved for temperature profile in my 3d comsol 4.2 model. Now how can i define the elements which are greater than melting temperature as fluid,can i export the data to matlab then identify the nodal points and then define separate physical properties for molten metal.

Earlier reply will be very useful to our progress.

Thanku in advance.

4 Replies Last Post Jun 17, 2016, 11:31 AM EDT
Posted: 6 years ago Oct 5, 2011, 12:21 AM EDT
Hi,

If i understood right, I think you need not to export file to define properties.

Is that for phase change and fluid flow?

First, you can use effective heat capacity method directly in comsol for melting

then,

add laminar flow study to marangoni effect.

Add carman-kozeny approximation in volume force term and weak contribution for surface tension effect

best regrads.

Hi, If i understood right, I think you need not to export file to define properties. Is that for phase change and fluid flow? First, you can use effective heat capacity method directly in comsol for melting then, add laminar flow study to marangoni effect. Add carman-kozeny approximation in volume force term and weak contribution for surface tension effect best regrads.

Kartikeyan Ramachandran
Posted: 6 years ago Oct 5, 2011, 4:54 AM EDT
sir,

thank you for your reply

the problem i am facing is how can i select only a particular nodes(which are greater than the melting temperature) as fluid and solve for the fluid equations in that domain alone... can i define the whole geometry as fluid(assuming solid to be a high viscous fluid) and then solve for heat transfer and fluid flow equations.

Earlier reply will be very useful 4 our progress

Thank u in advance

Karti Keyan
sir, thank you for your reply the problem i am facing is how can i select only a particular nodes(which are greater than the melting temperature) as fluid and solve for the fluid equations in that domain alone... can i define the whole geometry as fluid(assuming solid to be a high viscous fluid) and then solve for heat transfer and fluid flow equations. Earlier reply will be very useful 4 our progress Thank u in advance Karti Keyan

Posted: 6 years ago Oct 5, 2011, 9:56 AM EDT
You mean

"Fluid flow in only molten region"

right?

The whole region is considered as liquid.

Firstly, you should solve phase transition with "heat transfer module" by effective heat capacity method

In that way, you can define liquid volume fraction which range from 0 to 1. (1 for liquid, 0 for solid)

Then, you want to see the velocity profile only in molten region( fl=1), right?

Therefore you need to "porosity method" with "laminar flow module"

You can easily find the paper related to porosity method in google,

in brief, porosity=0 for solid, porosity=1 for liquid.

It can be defined in volume force term in "laminar flow module"


Fx = -C*((1-fl)^2)/(fl^3+b)*u
Fy = -C*((1-fl)^2)/(fl^3+b)*v+rho*g*beta*(T-Tinf)

fl : liquid volume fraction
C : sufficiently large number (ex) 1e12
b : constant for not divided by zero (ex) 1e-3
rho : density
g : gravity


This equation makes velocity zero in solid.

best regards



You mean "Fluid flow in only molten region" right? The whole region is considered as liquid. Firstly, you should solve phase transition with "heat transfer module" by effective heat capacity method In that way, you can define liquid volume fraction which range from 0 to 1. (1 for liquid, 0 for solid) Then, you want to see the velocity profile only in molten region( fl=1), right? Therefore you need to "porosity method" with "laminar flow module" You can easily find the paper related to porosity method in google, in brief, porosity=0 for solid, porosity=1 for liquid. It can be defined in volume force term in "laminar flow module" Fx = -C*((1-fl)^2)/(fl^3+b)*u Fy = -C*((1-fl)^2)/(fl^3+b)*v+rho*g*beta*(T-Tinf) fl : liquid volume fraction C : sufficiently large number (ex) 1e12 b : constant for not divided by zero (ex) 1e-3 rho : density g : gravity This equation makes velocity zero in solid. best regards

Antoni Artinov
Posted: 1 year ago Jun 17, 2016, 11:31 AM EDT
Hi JiYoung Park ,

is there a possibility to contact you in pm. I just started my Phd and have to do a CFD Simulation of a laser welding process taking into account all effects you described above. Somehow does my program not converge for the laminar flow case, but works well with a k-e turbulent model. I am quite sure that all physics parameters and bc are ok, but have the feeling, that I have to optimize the solver (conditions). Will really appriciate your help!

Best regards from Germany
Antoni
Hi JiYoung Park , is there a possibility to contact you in pm. I just started my Phd and have to do a CFD Simulation of a laser welding process taking into account all effects you described above. Somehow does my program not converge for the laminar flow case, but works well with a k-e turbulent model. I am quite sure that all physics parameters and bc are ok, but have the feeling, that I have to optimize the solver (conditions). Will really appriciate your help! Best regards from Germany Antoni

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.