[FLUENT]Marangoni effect
 

Hello, I've built 2D model of laser spot welding in Ansys Fluent.
Recently I wanted to add Marangoni BC so I've enable "Marangoni Stress" in "momentum" tab and enter 0.00029 [N/m*K] (for 304 stainless steel). Tried also other values for different sulfur contents. Values of surface tension gradient was taken from "Surface tension of molten stainless steels under plasma conditions" T. MATSUMOTO, T. MISONO, H. FUJII, K. NOGI as derivative dy/dt of y= 0.29×T+1030(y - gamma ;-) ).
My problem is that liquid zone is enormous compared to case without Marangoni BC.
In attachment I put a picture of the case with one laser pulse (timp = 30ms, Pimp = 1315W, radius = 0.350e-3 m, tophat) without the Marangoni effect and three cases with different values ​​of gamma. Both cases differs only in the inclusion of that effect.
Have any of you met with such a problem or have an idea what could be causing this?
P.S
cases are axisymmetric with symmetry to the x axi so there is only one half on the pictures.

The Marangoni stress plays a vital role. If you specify it negative the stream lines would tend to move outward of the free surface. Try to see them in Tecplot. Please send me your email address so that I send you a paper in which stream lines are also shown.

how to implement temperature dependent surface tension coefficient in fluent ?

Have you tried using the DEFINE_PROPERTY macro in a UDF?

Dear sir,
i have not tried UDF to apply property. but i have noticed when i enter values for Marangoni stress, it has provision for constant values, rather than a udf. is it possible to use temperature dependent Surface tension temperature coefficient under marangoni stress using udf ?

kcgmech, a old post but the marangoni stress value I input into the wall boundary condition in Ansys is the stress gradient due to temperature. So the value considers the surface tension changes in my fluid due to temperature variations. I was able to find a reference for this value for my fluid in the literature.