My objective is to study the variation in temperature as the fluid is expanded across a 3D nozzle. And therefore, I cannot specify temperature at the outlet.
Do I need to create a separate domain to specify the ambient temperature and pressure ?
You have to select energy equation in Fluent in the solver option. Once you select energy equation in the solver, you will find option THERMAL in both inlet as well as outlet boundary condition. You can give appropriate temperature at the outlet.
The problem could be solved by setting the temperature at the flow inlet and specifying a heat loss coefficient and ambient temperature as boundary conditions on nozzle surface (an adiabatic condition on this surface could work as well). That should be enough information and the outlet temperature should be a result from simulation.
The boundary condition at the outlet is required back-flow temperature, it does not mean, the exact value of the temperature at the outlet. The outlet temperature is calculated by extrapolation.
The thermal boundary condition at the outlet is zero gradient condition (dT/dx = 0). You need to specify static pressure and backflow total temperature at the outlet boundary. If there is no backflow at the outlet, then static temperature at the faces of the outlet boundary cells will be calculated using the condition dT/dx = 0. If backflow occurs through some cell faces at the outlet boundary, then the backflow total temperature specified by you will be used to calculate the static temperature at those outlet boundary cell faces.
the back flow in boundary condition we defined will be used when we have reversed flow at that boundary, so the fluent uses this value when faces back flow in system . otherwise you not need to know the exact value of back flow temperature and as Mr Mohamad Faisal said, the back flow temperature is not outlet temperature necessarily