Hi,
I'm trying to define a Robin boundary condition like "alpha(T1-T)=kdT/dx" for a heat transfer problem (time dependent). My first try was defining convective heat flux q0=h(Text-T), here the equation says -n.q=q0, so I assume if I define convective heat flux, I'll get the boundary condition as what I need. But the result shows that at this boundary, the line graph for q0 and k*dT/dx are too much different.
I guess this is not the right way. Any other idea? Thanks a lot!
Jing