[dramezan]

Hi,
I am attempting to simulate a problem in Ansys Fluent that involves a cylindrical domain partially filled with a particle sample. A laser source positioned at the top of the cylinder emits radiation onto a section of the sample's surface, causing it to heat up and melt. The challenge lies in simulating the phase change of the sample and the corresponding surface deformation. To accurately model the laser beam as a heat source (heat flux), assuming the surface of the sample is a wall is not feasible.

To simulate the laser beam, I employed the Discrete Ordinates (DO) solver and assigned the -y direction to the beam. Despite refining the laser beam path and using a large number of \(\phi\) and \(\theta\) in the DO configuration (up to 25 \(\phi\) and \(\theta\)), the laser beam path appears smeared and does not accurately represent the intended direct path.

In a smaller cylindrical domain, I achieved a nearly direct laser beam using 20 \(\phi\) and \(\theta\) for the DO solver. However, this approach does not yield the same result in my main problem, where the cylinder is long.

I am seeking advice on overcoming this issue and simulating directional radiation in Ansys Fluent without assuming the receiving surface to be a wall. Any suggestions or insights on improving the accuracy of the laser beam path in the larger cylindrical domain would be greatly appreciated.