ANSYS and convection coefficients

[Tunalover]

Folks-
Does ANSYS require the user to supply convection coefficients for thermal problems? What I'm getting at is: why use such an advanced solution method (an FEA code) for a fluids problem if the user is forced to use such coarse estimates of boundary conditions (convection coefficient)? This would be like shooting a mosquito with a shotgun. If such an estimation must be made why not use a true CFD code which does not require convection coefficients. Why then use ANSYS for such a problem?



Tunalover

 

[Transient1]

You are right that a CFD code is the best way to go for complex flow/heat transfer. However, there are a good number of cases where an approximation of a heat transfer coefficient is an acceptable analysis method. Sometimes you need to estimate the effect of convection to know if your about where the design needs to be. For certain cases the hand calculations exist to determine a very accurate (within 5%) heat transfer coefficient as a function of temeprature. That's about the same accuracy as a CFD code.

Often I model the unit/heat pipe interface as a convection boundary condition since the performance of a heat pipe can be given as a conductance in W/m^2-K with a fixed reference temperature. I hope this answers your question.