Learning Forum

[Robin Lorenz]

Hello

I am doing a simulation of a laser hardening process with Ansys 2023R2. Until now i used the following code with success on a flat surface model, now i have to simulate on an uneven surface of a gear tooth and the code is dying on me. The goal being to target the surface on the top of a single tooth. 

I think the problem lies with the orientation of the code. Does somebody know a solution to this? When i start the simulation, the results show no thermal influece on the body of the gear, with maximal global temperature of the simulation showing only the general surrounding temperature of 22°C. 

The code surrounding the seeming error code serves the implementation of a lissajoux-path on to the laser, worked just fine on the flat surface model.

[dlooman]

You seem to be apply a heat generation rate where it would seem to be more typical to apply a surface heat flux.  With a heat generation rate the size of the element will determine how much heat you are applying.  The volume of the elements in the mesh you show will result in a big variation in the amount of heat.  The geometry calculations you do in the  *do loop seem too simple to apply to heat load moving over the surface of the gear teeth.  Could you explain how they are supposed to work?  If the laser is moving circumferentially at a constant rate along the surface as it goes up and down the gear teeth?

[Robin Lorenz]

 

Hi Dave

In the end, the Topheadlaser should move along a Lissajoux-Figure centered on the top of a single tooth. I have tried to position the laser by adding and subtracting from the centrx/centry/centrz commands without success. The velocity of the laser along the surface is linked to the power of the laser to combat temperature extremes at the edges of the lissajoux-figures. Do you know per chance what the marked *do-section would need to move along the surface of the gear? 

As the Lissajoux Figures only have an amplitude of about 1mm it should stay on the top of the tooth and not need to incorporate movement up and down the tooth. (Or so i thought)

Concering the heat generation rate, i tried to link it to a fixed radius, which i define at the beginning of the code as “rmax” to control the area where the heat is being applied. This somehow seemed to work in a previous project with no corelation to the element size. Do you this is gonna be a Problem in this geometry?

 

[dlooman]

So the laser is just heating the tip of the tooth?  Volumetric heat generation doesn't seem like an appropriate way to model the surface heating of a laser.  If you create a very fine mesh of the tooth tip area you could select nodes on the surface based on their theta coordinate in a cylindrical system.  You could update the theta values over time:

csys,1   ! global cylindrical 

nsel,s,ext  ! nodes on the surface

nsel,r,loc,x,mid_tooth,top_tooth+increment

nsel,r,loc,y,theta,theta+small_angle  ! needs to select at least two nodes for a face to be defined

sf,all,hflux,value

allsel

time,...

solve

reselect

next time 

solve

soo

[Robin Lorenz]

 

Thank you very much for the fast reply.

Yes, the laser just needs to dance on top of the tooth. I followed the tutorial by Steffen B. to create the original code.

(

&list=PLo1pvMoJK_x9_GefnNmUKcGU37sK04bQw)

I chose the volumetric heat generation to apply a 3-dimensional gauss structure to the laser, as that was one of the requirements i got for this project. Is there another way to achieve this? 

Is there a way to implement a function into this code structure to make it follow a Lissajoux-path? 

Please excuse me if any of my questions seem nonsensical or basic, they are ment genuinely. I am not very proficient in APDL and trying my hardest to complete this project, we do not have any experts on this kind of simulation in our institute. Learned Ansys Mechanical through online tutorials. 

 

[dlooman]

I see the demo model has a thin uniform mesh so that the heat input can be quantified.  I still think a surface flux (SF) makes more sense.  An ambitious first step would be to just get a narrow uniform heat flux to move across the tip of the tooth using the method I showed.

[Robin Lorenz]

Thank you very much for your assistance. 

Could a surface flux result in the same results as a Gaussian Laserprofile?

Or could it be possible if i mesh the target region as small and evenly as possible and target specific nodes with the code in order to reuse the other parts of the script? I was able to validate the results of the code in practice on a simple geometry via metallurgical analysis.

 

 

[dlooman]

Yes, I believe so.