SimScale CAE Forum

'Thermal Analysis of a Laser Beam Welding ' simulation project by ahmedhussain18


#1

I created a new simulation project called 'Thermal Analysis of a Laser Beam Welding ':

In this project a thermal analysis of a laser beam welding (LBW) was performed.


More of my public projects can be found here.


#2

Laser beam welding (LBW) is a welding technique normally used to join pieces of metal by using laser beam. The welding is performed by indenting a concentrated heat source provided by a laser beam over the area that needs to be welded. In this project thermal analysis of a LBW was demonstrated. The main focus was on the use of surface heat flux boundary condition to produce the effect of moving heat source. The geometry was a simple circular plate on which a small hump in a circular fashion was made to represent the welded area. The geometry is shown in the figure below.

The geometry was meshed using tetrahedralization with refinement on SimScale platform with much finer mesh over the welding area. The mesh is shown in figure below.

A transient thermal analysis was chosen as analysis type. Steel was selected as a plate material. All the surfaces of the plate except the bottom surface were considered exposed to air by giving the convective coefficient value of 5 W/m2 K at room temperature i.e. 298.15 K (25 oC). In order to have a moving heat source in circular fashion, the welding area was given a surface heat flux by using a Gaussian power distribution function represented as:

  • P is the power of laser, 2 kW in our case.

  • ro is the radius of laser beam, 0.003 m (3 mm) in our case.

  • r = √ (x-xo)2+(y-yo)2 is the radial distance of the beam from any point. Here, x and y are x and y direction of the coordinate system respectively. xo and yo are the terms used for defining the motion. In our case,
    xo = a + R cos(2π vα t) and yo = b + R sin(2π vα t), where;

  • a and b are the x and y coordinates of the circle center, in our case (a,b)=(0,0).

  • R is the radius of the circle, 0.035 m (3.5 cm) in our case.

  • vα is the angular velocity of the beam, 0.167 rad/s in our case. The actual velocity of the beam can be calculated by v = 2πRvα which gives v ~ 0.037 m/s (~2.22 m/min.)

  • t is time step which is 6 s for a full revolution with timestep length of 0.06 s in our case i.e. cos(2π) and sin(2π) occurs when vα t = 0.167 rad/s x 6 s = 1 rad

The analysis was run on 8 cores and took around 16 minutes to finish. The figures below show the contour plots of temperature and heat flux of a circulating laser beam respectively.

Finally, the figure below shows the circulating laser animation.

Note:

Although the temperature of the laser beam was much higher than the highest temperature of the legend i.e. around 2500 K, the trimmed value of 1783.15 K is used since it represents the melting temperature of steel.


#3

I need some help to simulate a laser beam heat a teeth superficie. Can you help me?


#5

Sure! how can I help you?


#6


#7

Can you help me to create a thermal model for laser transformation hardening of helical gear..I have the model in IGES format.


#8

Hello,


I need the same moving heat source in terms of laser cladder.(Laser build up welding)

Could you please help me out


Thank you


#9

Hi @sseshadri!

If you upload your CAD model we can give it a try.

Best,

Jousef


#10

This is how it looks like
I am not able to move the heat source at a point of time and depending upon the clad length…

It would be helpful if you coulc guide me in this…


#11

Hi @sseshadri! All you need to do is to replace xo and yo in:

with a center point value of your laser beam + velocity. To have an idea, please see this project: Helical Gear Laser Hardening Simulation

Your ‘r²’ can be something like: ((x-(xo* t))+(y-yo))


#12

Thanks for your answer but do you have any reference document for the calculations???


#13

here you go: https://www.ajol.info/index.php/ijest/article/viewFile/59285/47583


#14

thanks alot…:slight_smile:


#15

Have you come across doing same simulation process in Nx nastran??


#16

Nops… Sorry!


#17

Very good write up! Have you thought about simulating friction welds? DO you think it would be difficult?