Discussion Forum

Hi everyone,

Firstly, happy Thanksgiving Day and thank you for your response to my post. I’m trying to build a laser-matterial interaction model based on the “heat transfer mode” and “moving mesh mode”. During those days, I have encountered some problems. Could you give me some guidance? Thank in advance for your response.

Model details:
The model geometry is a simple rectangle (2D). A laser beam is treated as a heat flux boundary on 4th (boundary number or the upper boundary) boundary (the heat flux is limited in a certain range-narrow than the length of the 4th boundary), while the other three boundaries are “thermal insulation”.

Ps. all the details are illustrated in the attachment. The numbers with circle means boundary, while the one without circle is domain. The ALE (attribution - prescribed mesh displacement) is applied to the upper boundary (number 4), while other is fixed. Thanks.

Physical basics:
When the material temperature is getting higher than T_e (usually equal to the evaporation temperature or melting point), this part of material will be removed, so the new boundary is created. Few papers have used ALE to trace the moving interface successfully, so I want to have a try.

Idea and unsolved problem:
Once before, I had no idea about how to deal with the prescribed mesh displacement. Recently, another idea comes to my mind. Franking speaking, I don’t know whether it works satisfactory/well or not, so could you help me check it? Thanks.
The detailed idea:
1. When the temperature is higher than T_e.
2. Extract the temperature at each nodal point; assuming the coordinate of the old nodal point is (x, y).
3. Find the new nodal point (the blue point in the attachment photo), where the temperature is just equal to T_e, mabye using an interpolation funciton or others; assuming the new coordinate is (x, y’) at a specific x (identical to the x value of the old nodal point for a good compare and calculation).
4. Define a variable ”dis_y”.which is the substraction of the x value between new and old nodal points. i.e. dis_y=y-y’
5. Using dis_y to present the prescribed mesh displacement in the ALE.
6. The COMSOL moves the upper boundary according to the prescribed mesh displacement “dis_y”. Furthermore the new boundary is created like the blue dashed line in the attachment.

The aim of the second and third one is to find the coordinate where the temperature is identical with T_e. could it be achieved or feasible? About this, can using a “for” statement (to select coordinates in the geometry area whole) and “mphinterp or some evaluation funciton” in LiveLink with Matlab handle it?

Happy Thanksgiving Day and best wishes to you and your family.

Yours
FM Huang
Nov. 24th .2011