Here is an example from K&W chapter 12  The heat flow coefficient Kp for turbulent flow in smooth passages (radiator core passages)   Kp = 1/2*L/D* 0.326/(Re)^-4  . 

 

  If Kp is a measure of goodness, then clearly if L increases and D gets smaller Kp increases.  Or if the Reynolds number Re gets smaller Kp goes up.  So what does this mean?  It basically shows that for the heat transfer to be large, the Reynolds number should be low (I.e. the airflow through the core should be slow), the core should be deep(large L) and the hole's hydraulic diameter (D) should be small. 

 

 This makes sense as the thicker the core the more heat transfer (although the further into the core the less efficient the heat transfer), the holes should be smaller (area exposed area - with large holes some of the cooling air in the center will simply not have as much contact with the hot metal of the core)  and the air velocity should be slow (dwell time adds to heat transferred to the unit volume of air per unit time). 

 

However, if you make the core too thick or the holes too small or slow the air too much -  then your KP factor may be high - but your over all cooling will suck because you have too little mass flow through a too restrictive core.  This is just one example of where optimizing on one set of factors can play havoc with the overall system function.   One way of looking at it is that you have to suboptimize a lot of factors in order to get an optimum system {:>)

 

My 0.02

 

Ed