 ## Fluids

• gao.1571
Subscriber

Hi, in my case the surface tension gradient is dependent on the temperature and a user-defined-scalar. i model this in two ways,

1. the first one is to use the built-in Marangoni stress in Wall Momentum panel. Use an expression of temperature and UDSI(0) to define the surface tension gradient. But the expression is not single-valued. A example is like this, 2.03*log(1.0+ UDS(uds = 'uds-0') * exp(StaticTemperature). So here the expression is limited to a single valued quantity?
2. the second is to use udf by coding the Marangoni stress = d(gamma)/d(T)*d(T)/d(x). I get the d(T)/d(x) by the marco C_T_G. This method gives a much larger velocity than the built-in method in a test case (20 m/s vs. 1 m/s). in the test case, the d(gamma)/d(T) = -5e-4. i suspect the built-in uses a differnt temperature gradient than that obtained by C_T_G.

I appreciate all the suggestions.

• DrAmine
Ansys Employee

How are calculating the stress and which results in which macros are you returning back to Fluent?

• gao.1571
Subscriber

Hi,

1. for the built-in Marangoni stress, i input value in this panel. i want to ask if the surface tension gradient can be a function of Static temperature or it must be a constant? 2. for the udf method, i use the DEFINE_PROFILE to set the shear stress boundary condition. In it, the marco C_T_G is used to get the temperature gradient. i want to ask this temperature gradient is the same as that used by the built-in Marangoni model. Because i find the velocity of the udf method is like 20 times larger than that of built-in method.

• DrAmine
Ansys Employee

1) Single valued field: so you need to return back a constant value field.

2) I think it should be the same temperature gradient. If the results are completely off then either your UDF is wrong or the physics you are implementing in the UDF are having some logic issue.

• gao.1571
Subscriber

1. But i don't understand why it provides a window for various definitions when the constant value is the only choice. it seems it can be a function of variables, such as the static temperature. Can you double-check that? 2. Can you provide the theory basis to implement the marangoni stress for a single phase problem?

• DrAmine
Ansys Employee

That Window is not only for that field. We have a multi-purpose code. If you try to provide a profile for that field you cannot do that using UDF for example.

Can you tell me which correlation are you trying to follow or which formula you want to assess? Do you want to give temperature based conditional for the treatment?

The theory is mentioned in the documentation: you require surface temperature gradient.

• gao.1571
Subscriber

Sorry i don't understand this sentence: "If you try to provide a profile for that field you cannot do that using UDF for example." You suggest me to use the udf to input a profile? or the profile is invalid for that field.

i want the surface tension gradient as a function of temperature not a constant. gamma = F(temperature).

Thanks.

• DrAmine
Ansys Employee

Even using DEFINE_PROFILE to provide that coefficient is not possible.

I think you require to provide the shrea stress than as the theory guide is describing for marangoni stress.

I will think bit more about that if there is another simpler way and update this the next days/week. 