It has been estimated that, of all the worldwide energy utilization, more than 80% involves the heat transfer process, and the thermal engineering has for a long time recognized the huge potential for conserving energy and decreasing CO2 release so as to reduce the global warming effect through heat transfer efficiency techniques (Bergles, 1988, 1997; Webb, 1994; Zimparov, 2002). In addition, since the birth of electronic technology, electricity-generated heat in electronic devices has frequently posed as a serious problem (Arden, 2002; Chein & Huang, 2004), and effective cooling techniques are hence needed for reliable electronic device operation and an increased device lifespan. In general, approaches for heat transfer enhancement have been explored and employed over the full scope of energy generation, conversion, consumption and conservation. Design considerations to optimize heat transfer have often been taken as the key for better energy utilization and have been evolving into a well-developed knowledge branch in both physics and engineering. During the last several decades and promoted by the worldwide energy shortage, a large number of heat transfer enhancement technologies have been developed, and they have successfully cut down not only the energy consumption, but also the cost of equipment itself. However, comparing with other scientific issues, engineering heat transfer is still considered to be an experimental problem and most approaches developed are empirical or semi-empirical with no adequate theoretical base (Gu et al., 1990). For instance, for a given set of constraints, it is nearly impossible to design a heat-exchanger rig with the optimal heat transfer performance so as to minimize the energy consumption. Therefore, scientists developed several different theories and methods to optimize heat transfer, such as the constructal theory (Bejan, 1997) and the minimum entropy generation (Bejan, 1982). Then heat transfer processes were optimized with the objective of minimum entropy generation. Based on this method, several researchers (Nag & Mukherjee, 1987; Sahin, 1996; Sekulic et al., 1997; Demirel, 2000; Sara et al., 2001; Ko, 2006) analyzed the influences of geometrical, thermal and flow boundary conditions on the entropy generation in various convective heat transfer processes, and then optimized them based on the premise that the minimum entropy generation will lead to the most efficient heat transfer performance. However, there are some scholars (Hesselgreaves, 2000; Shah & Skiepko, 2004; Bertola & Cafaro, 2008) who questioned whether the entropy generation is the universal irreversibility measurement for heat transfer, or the minimum entropy generation is the general optimization criterion for all heat transfer processes, regardless of the nature of the