It is important to appreciate that it may not be possible to attain the design point simply by heating the evaporator or cooling the condenser; the heat pipe may not "start."

Cotter (1967) uses the terms "uniform start-up" and "front start-up". In uniform start-up the temperature along the heat pipe is practically uniform during the start-up process; this is obtained where initially the vapor density is not too low.

In frontal start-up there is a considerable temperature variation along the pipe, and it is only gradually that a relatively uniform axial temperature variation is obtained.

When the density of the vapor is initially very low, it is possible for some flow to occur. This will start first at the outlet from the evaporator, or at the outlet from the isothermal length. Cotter's treatment has now been found to be inadequate in the light of work with lithium heat pipes carried out by Sockol and Forman (1970). They present the equations for predicting the rate at which the sonic front will move along the heat pipe to the condenser.

Downstream of the sonic front, supersonic flow may occur, followed by a shock wave.