Menu

Discussion Forum

New Discussion
Filter

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 The information provided may be out of date.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.

Element killing during laser melting

Posted Jul 18, 2013, 10:40 a.m. EDT Heat Transfer & Phase Change, Materials, Mesh, Modeling Tools & Definitions, Parameters, Variables, & Functions, Structural Mechanics Version 4.1, Version 4.2, Version 4.4 7 Replies

uday
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Dear friends

I am working on laser cutting with moving heat source with subsequent stress analysis. I have successfully simulated the thermal analysis . However, I dont know how to delete/kill those elements whose temperature goes beyond the melting temperature during laser heating. for example, Assume that the melting temperature is 1273 K, thus any element with a temperature of 1273 or greater must be killed to simulate melting.
comsol can do this element kill process?
Can anybody guide me in this issue?

Thanks in advance

7 Replies Last Post Jun 7, 2017, 4:25 p.m. EDT
Josh Thomas Certified Consultant
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Jul 18, 2013, 2:03 p.m. EDT
Uday,

One way to do it is to make the Young's Modulus and Poisson's ratio functions of temperature and use a ramp function to ramp the values as dependent on Temperature across the phase change.

As the temperature increases across the melting point:

1) Drop the Young's modulus to zero (or near zero to avoid convergence difficulties).

2) Raise the Poisson's ratio to 0.5 for incompressibility (or near 0.5 to avoid convergence difficulties).

So the elements are not "dying" but they are contributing effectively zero stiffness once they are "melted".

There may be other ways of doing this, but this is one simple way.

Best regards,
Josh Thomas
AltaSim Technologies
Uday, One way to do it is to make the Young's Modulus and Poisson's ratio functions of temperature and use a ramp function to ramp the values as dependent on Temperature across the phase change. As the temperature increases across the melting point: 1) Drop the Young's modulus to zero (or near zero to avoid convergence difficulties). 2) Raise the Poisson's ratio to 0.5 for incompressibility (or near 0.5 to avoid convergence difficulties). So the elements are not "dying" but they are contributing effectively zero stiffness once they are "melted". There may be other ways of doing this, but this is one simple way. Best regards, Josh Thomas AltaSim Technologies
uday
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Jul 19, 2013, 7:35 a.m. EDT
Hi Thomas,

Thaks for your valuable reply. But , still i having trouble over there.

I could not understand your point. can you give some idea

1) Drop the Young's modulus to zero (or near zero to avoid convergence difficulties).

2) Raise the Poisson's ratio to 0.5 for incompressibility (or near 0.5 to avoid convergence difficulties).


can you give me some explanation about your points

thanks
Hi Thomas, Thaks for your valuable reply. But , still i having trouble over there. I could not understand your point. can you give some idea 1) Drop the Young's modulus to zero (or near zero to avoid convergence difficulties). 2) Raise the Poisson's ratio to 0.5 for incompressibility (or near 0.5 to avoid convergence difficulties). can you give me some explanation about your points thanks
Josh Thomas Certified Consultant
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Jul 19, 2013, 9:03 a.m. EDT
You need to make the material properties (E and nu) functions of Temperature.

These functions will "ramp" from the values of E and nu for the solid phase to the values of E and nu for the liquid phase right at the melting temperature. The value of E for a liquid is zero for E and 0.5 for nu. Don't use these values exactly, but rather use some value very close to them. This will help with solver convergence.

Also, when solving for materials with values of nu close to 0.5, under Solid Mechanics>Linear Elastic Material you should check "Nearly incompressible material" -- this adds an addition degree of freedom for the pressure but the mixed formulation is better for these materials.

That's the best I know to explain it!

Best regards,
Josh Thomas
AltaSim Technologies
You need to make the material properties (E and nu) functions of Temperature. These functions will "ramp" from the values of E and nu for the solid phase to the values of E and nu for the liquid phase right at the melting temperature. The value of E for a liquid is zero for E and 0.5 for nu. Don't use these values exactly, but rather use some value very close to them. This will help with solver convergence. Also, when solving for materials with values of nu close to 0.5, under Solid Mechanics>Linear Elastic Material you should check "Nearly incompressible material" -- this adds an addition degree of freedom for the pressure but the mixed formulation is better for these materials. That's the best I know to explain it! Best regards, Josh Thomas AltaSim Technologies
Tianhang Bai
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 9 years ago Feb 25, 2015, 12:03 a.m. EST
Hi Josh,

I tried to reduce the Young's modulus to a lower value, but it did not converge even if the change is small. Do you know how to fix that? Thanks.

Regards,
Tianhang
Hi Josh, I tried to reduce the Young's modulus to a lower value, but it did not converge even if the change is small. Do you know how to fix that? Thanks. Regards, Tianhang
Josh Thomas Certified Consultant
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 9 years ago Feb 25, 2015, 9:00 a.m. EST
You could continue to increase E_liquid by an order of magnitude at a time until convergence. As long as E_liquid is still much much less than E_solid you should get meaningful results.

--
Best regards,
Josh Thomas
AltaSim Technologies
You could continue to increase E_liquid by an order of magnitude at a time until convergence. As long as E_liquid is still much much less than E_solid you should get meaningful results. -- Best regards, Josh Thomas AltaSim Technologies
Shubham Shah
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 9 years ago Jun 27, 2015, 5:41 a.m. EDT
But the issue is once the element has melted, it should absorb latent heat of melting before melting and its temperature should not increase after melting.
But the issue is once the element has melted, it should absorb latent heat of melting before melting and its temperature should not increase after melting.
Antoni Artinov
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 7 years ago Jun 7, 2017, 4:25 p.m. EDT
Hi Josh,

just to manipulate the Young's modulus is not enough ( to set the Poisson's ratio to about 0.5 will have really hard numerical consequences, so to be aware with that!). After melting of the material the element's strain should be set to zero so that the material history will be removed and the strain starts again from zero by the solidification of the part. Unfortunately, there is still not a feature in COMSOL Multiphysics which allows this and as much as I know also not a proper workaround.

Best regards,

Antoni
Hi Josh, just to manipulate the Young's modulus is not enough ( to set the Poisson's ratio to about 0.5 will have really hard numerical consequences, so to be aware with that!). After melting of the material the element's strain should be set to zero so that the material history will be removed and the strain starts again from zero by the solidification of the part. Unfortunately, there is still not a feature in COMSOL Multiphysics which allows this and as much as I know also not a proper workaround. Best regards, 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.

Suggested Content