EXPERIMENTAL STUDY OF HEAT LEAKAGE. 779 



varies more nearly as the pressure drop — is, in fact, in practically 

 every case, proportional to pressure drop, within the limits of experi- 

 mental error. Since the temperature drop increases more rapidly 

 than the pressure drop in all cases, this means that the heat leak per 

 unit time increases less rapidly than the temperature drop. 



The very large leakage in plugs Al, A2 and A3 is partly due to the 

 unusually happy combination for the promotion of heat-leak afforded 

 by the long and rather constricted cross-channel used in these plugs. 

 The comparatively small area of cross-section of the cross-channel, by 

 forcing a high steam velocity and thus increasing the kinetic energy 

 of the steam at the expense of its internal energy, depresses the 

 temperature of the steam while it is in the cross-channel by an amount 

 which, for high velocities, varies nearly as the inverse fourth power of 

 the diameter of the cross-channel, for a given flow, and as the square 

 of the flow for a fixed cross-section of the cross-channel. Since for 

 these plugs the flow is approximately proportional to the square root 

 of the pressure-drop, the temperature depression due to kinetic energy 

 effect alone in the cross-channel will vary directly as the pressure 

 drop. To this must be added the depression due to the throttling 

 which has taken place in the high-side half of the plug, which is also 

 proportional to pressure drop. Of course this total cross-channel 

 depression of temperature is, in the steady state, reduced to some 

 extent by the heat-leakage itself. It might be expected that the 

 kinetic energy effect would predominate. This is not borne out by 

 the facts, however; for plug A2, which permits more than 50 per cent, 

 greater flow for a given pressure drop than plug Al (see Table II) 

 would produce a kinetic energy temperature depression more than 

 twice as great as plug Al for a fixed pressure drop, while the throttling 

 depression would be the same for both. If then the kinetic energy 

 effect were the dominant one, the heat leak per gram of steam at a 

 fixed pressure drop might be expected to be greater for plug A2. 

 The contrary is the case, as is shown by the A T vs. Ap plot. The heat- 

 leak per gram of steam per unit pressure drop for a given flow is, how- 

 ever, greater for plug A2, as is indicated by the ji' vs. - plot. This is 



because the pressure drop for a given flow is smaller for A2 than 

 for Al. 



The A 7' vs. Ajj plots for these three plugs are all concave upward, 

 though only slightly so; this means that the heat leak per gram of 

 steam, which increases fairly rapidly with the pressure drop through- 

 out the range of the experiments for all three plugs, increases a little 



